Telecommunications, Media, and Technology

RECALL Noxx No13

Leveraging Technology Title

RECALL No 13 Leveraging Technology

Welcome ...
... to RECALL No 13. For the first time, we are issuing an electronic version of RECALL, our publication for leaders in the telecommunications industry. Paper copies are available on request. The individual articles can also be found in McKinseys Telecommunications Extranet. The current issue focuses on the huge potential that still remains to be captured in the IT and technology arenas. Our first article synthesizes the results of our international Telecoms IT Benchmark surveys most recent wave and dispels a number of myths about whether and how IT and performance are correlated. We then show the structured application of lean techniques that can deliver significant performance improvements across key business areas. We start by illustrating how telcos can unlock significant productivity gains in their field force by using smart simulation across their entire operating model. Next, we discuss how IT-enabled lean transformation can boost performance in application development and maintenance along all aspects of the organization. In another dimension of lean, we look into the service delivery chain for B2B information and communication technologies, where lean can tap productivity improvements of some 40 percent. The fifth article describes a deeper, more collaborative approach to outsourcing and offshoring that can achieve radical transformation of an operators IT landscape. Such an approach also makes it possible to realize high incremental cost savings. We then move on to nextgeneration IT architecture management, revealing the best practices of top performers. Our final article drills down from this broad view to a specific, yet vital facet: customer lifecycle management. Here, the right infrastructure can greatly reduce churn and lift revenues. In closing, we hear from Fernanda Torquati, Global CIO of Telefnica. She shares with us the unique story of her companys IT transformation journey in an interview. She believes they will achieve a uniform global IT at the end of this journey, while ensuring local demand management and an overarching country interface. But such journeys dont happen overnight. We hope this issue of RECALL sparks insights and discussion as you navigate your own technology journey. As always, we look forward to your feedback on these articles, as well as your thoughts on topics you would like to see covered in the future.

Jrgen Meffert Leader of McKinseys EMEA TMT Practice

Tomas Calleja Co-leader of McKinseys Operations and Technology in TMT Practice

Klemens Hjartar Co-leader of McKinseys Operations and Technology in TMT Practice

Fabian Blank Leader of McKinseys EMEA Mobile Operations Service Line and Editor of this RECALL issue

RECALL No 13 Leveraging Technology

Contents
01 02 03 04 05 06 07 08 Money or management? The true driver of telco IT performance Sweating your assets: Riding the next wave of savings in the field force Lean on the line: Improving ADM in telecoms Fixing break-fix: The power of lean ICT transformation Outside in: Leveraging outsourcing and offshoring to transform IT Fast forward to success: Managing next-generation IT architecture Turning customer insights into income: The architecture is key Transforming IT: An interview with Fernanda Torquati, Global CIO, Telefnica 7 13 19 25 31 37 45 53 57

Appendix

RECALL No 13 Leveraging Technology Money or management? The true driver of telco IT performance

01 Money or management? The true driver

of telco IT performance

Conventional wisdom dictates that greater IT spend leads to improved performance. Results of McKinseys Telecoms IT Benchmark, however, show that the secret to IT efficiency and effectiveness lies in its management. Since 2008, McKinsey & Company has been conducting an IT benchmarking effort in telecoms. The exercise not only aims to measure and compare IT efficiency (i.e., how much do I spend on IT?) and effectiveness (i.e., how well does IT support my company?) but, most importantly, to identify key performance drivers. The survey has highlighted substantial differences across participants in these efficiency and effectiveness categories. McKinseys IT effectiveness index defines performance along three dimensions that are crucial for every telecoms operator: Time to market for new products/services Functionality coverage by business activity Availability of business-critical IT services. The higher the score, the better the telcos IT supports the organization. Comparative analysis of results emphasizes time to market as the dimension that explains most of the difference between best and poorest operators. This is highly significant, since top performers can be up to five times faster in bringing new customer price plans to the market. IT efficiency is simply the ratio of IT spend (capex plus opex) to company revenues. The two years of this study

have revealed that what typically differentiates more efficient from less efficient operators is their spending in application development. With the IT benchmark, McKinsey has been able to identify the areas that drive excellence in IT management (i.e., the key management practices that distinguish top performers) and to dispel some myths, while highlighting several realities.

Drivers of excellence in IT management

In this study, McKinsey examined IT management practices of operators and identified the factors that differentiate the most efficient and effective operators from the least. The best operators excel mainly in the following three areas: Demand management. Top telecoms performers strictly adhere to demand management processes, and this holds true both for application development and for application maintenance. Some examples of best-inclass demand management are (Exhibit 1): Deploying and adhering to a stringent, gated process to review the demand, facilitating focus and accountability of stakeholders, and establishing a more efficient and effective decision making process Creating a yearly IT capacity plan by segmenting the demand along two dimensions business lines and demand type (i.e., projects, small demands, and maintenance)

01

The best operators adhere strictly to formal demand management processes The best operators adhere strictly to formal demand management processes
Operators lagging behind Applications going through formal demand management process Percent Basic/low Question 100 Development 76 To what extent are account managers vs. the business involved in demand management? What is the involvement of the business in approval of IT projects? Is there an infrastructure product catalog (excluding storage)? How is the chargeback mechanism used? Best operators Advanced/high

100 Maintenance 87

SOURCE: McKinsey 2009 European Telecoms IT Benchmark Survey

MOBILE OPERATORS

Appointing a set of sufficiently senior and knowledgeable relationship managers from IT to the business to help steer their business counterparts toward the optimal functionality cost versus timeto-market trade-off. Top performers are also more advanced in having defined an infrastructure service catalog. This allows

them both to standardize the factory (in terms of process, services, technologies, etc.) and to have the correct demand management dialog with users. These activities ensure that the service level provides adequate quality at the right cost. Vendor management. IT purchasing represents 60 to 70 percent of the total IT spend for telecoms operators. This means that vendor management is a top priority. The first lever for successful vendor management is vendor consolidation. McKinseys Telecoms IT Benchmark shows that top performers have far fewer overall IT vendors than do the poorest performers. Top fixed-line performers, for example, have 60 vendors, while operators coming in at the bottom of this category have 133. It is not unusual, in fact, to find the following situation at a telco: a plethora of both large and small vendors in application development and maintenance (ADM) and infrastructure, multiple contract models, limited service quality targets, no vendor risk of losing business continuity, limited application of contract penalties, lack of cost transparency mechanisms, and hundreds of RFPs to purchase IT products and services. IT vendor consolidation represents a formidable opportunity to achieve a substantial reduction in IT spend and a significant improvement in service quality. Top

McKinseys Telecoms IT Benchmark

In 2008, McKinsey conducted its first IT benchmarking study in telecoms with a sample of 19 European operators, including 6 integrated players, 3 fixed-only operators, and 10 mobile-only operators. The 2009 survey has seen participation grow to a total of 27 participants in Europe and the first set of benchmarks in North America. The study participants are typically fixed incumbents and large mobile players (i.e., no mobile virtual network operators). The benchmark measures both IT spend as a proportion of revenue as well as IT effectiveness to put IT costs in the context of service levels, time to market, and the functionality that IT systems offer to the business.

RECALL No 13 Leveraging Technology Money or management? The true driver of telco IT performance

02

More outsourcing does not necessarily mean fewer in-house FTEs More outsourcing does not necessarily mean fewer in-house FTEs
FTEs1 per EUR billion in revenues 900 800 700 600 500 400 300 200 100 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Percent of outsourcing
1 Including internal onshore and offshore plus staff contractors SOURCE: McKinsey 2009 European Telecoms IT Benchmark Survey

Fixed

Mobile

R2 = 0.21

APPLICATION DEVELOPMENT AND MAINTENANCE

performers are able to reduce IT purchasing spend by up to 25 percent (15 to 20 percent of total IT expenditure) by transforming their vendor management model. Operators should, however, be careful that they do not become too dependent on one service provider, especially for ADM. With only one provider, what were once benefits could turn into liabilities, since the operator becomes locked in. This is why companies should pursue vendor consolidation while maintaining a multisourcing strategy. The second lever of vendor management is a telcos outsourcing strategy. Outsourcing application development is often seen as an easy means of reducing ADM costs, but operators that outsource more dont necessarily have fewer non-outsourced staff (Exhibit 2). McKinseys Telecoms IT Benchmark has clearly highlighted that this is true only if the operator has mature vendor management practices, such as using detailed service definitions during negotiation, frequent contract reviews, time-to-market KPIs, and active vendor cost management. IT architecture. The third area of IT management is architecture, and McKinseys benchmarking study shows that top performers have a much more consolidated application landscape. As an example, operators at the top of the mobile list have an average of around 100 appli-

cations, while the poorest performers have about 170 (Exhibit 3). Even if application does not always mean the same thing across companies, these figures clearly indicate that a lower number of applications is a key performance driver in IT both in cost and effectiveness. Top performers also adhere much more to standard package functionality to avoid falling into the common trap of adapting them too much, thus turning them into another in-house development legacy. Fixed-line operators are much less advanced in using standard packages, mainly due to the large legacies in this particular business. As another example of best-in-class architecture, a high consolidation level of corporate databases (product catalog, installed base, etc.) enables operators to reduce ADM spend and improve time to market. Some operators, for example, have run projects to build an integrated product catalog and to implement new product development through simple parameterization. This allows them to drive down the time to deploy new price plans to as little as just a few days.

The truth about IT in telecoms

McKinseys Telecoms IT Benchmark also helped dispel some myths and highlight several of the realities of how IT supports telecoms operators. The key findings from our survey are the following:

10

03

The best operators have simpler application landscapes The best operators have simpler application landscapes
Business support systems Operations support systems Enterprise resource planning Mobile Number of applications Best operators 16 16 Fixed Number of applications

65

97

141

85

50

276

Overall

85

29

30 144

166

105

49

320

Poorest operators

103

39

26 168

172

145

90

407

SOURCE: McKinsey 2009 European Telecoms IT Benchmark Survey

Telecoms operators can deliver high IT effectiveness with limited IT spend, i.e., 3 to 4 percent of their revenues. The myth of needing to spend more to raise IT effectiveness has been refuted. What really moves the needle is management practices, and this holds true for both mobile and fixed-line operators as well as in other industries.

Size does not necessarily help. Small operators can be efficient and the bigger ones too. But we believe that large operators are not fully capturing the benefits of being large, resulting in some missed opportunities. This situation is often driven by the excessive complexity of the IT landscape. High levels of outsourcing do not necessarily imply low IT spend. In fact, operators whose IT spend is a large percentage of their revenues often outsource a great deal of their IT functions without bringing their IT spend to below-average levels at least in the short term. As outlined before, to capture the potential of outsourcing and offshoring, telcos must have mature capabilities in these areas. Standardization of technologies and interfaces improves time to market. The focus of time-to-market strategies tends to be on outsourcing, but poor outsourcing management can actually worsen the situation. Telcos who focus on IT standardization are the ones who see positive results in their time-to-market improvement efforts. Top performers are characterized by fewer, more productive human resources. This is not to say that a reduction in headcount improves performance. It does appear, however, that telcos that have been able to elimi-

A cross-industry perspective
While McKinseys Telecoms IT Benchmark offers insights within the telecoms industry, our work with other sectors confirms these learnings. In particular, top telco managers especially CIOs often ask how their organizations compare with banks in terms of IT spend and IT effectiveness. McKinseys telecoms and banking benchmarks indicate that contrary to popular belief banks are not much more advanced than telcos, even though they spend much more on IT (i.e., 7.5 percent of revenues versus 4 percent on average in telcos). Banks application development practices are also no more advanced than those of telcos, even though banks in general do have better performance management systems (e.g., using function points much more frequently).

RECALL No 13 Leveraging Technology Money or management? The true driver of telco IT performance

11

nate unnecessary complexity have smaller but more productive workforces. An overview of telecoms operators across type and over time confirms that outdated thinking regarding investment in and the handling of IT is not serving telcos well. Among others, the myths of more money equals better performance and greater outsourcing means lower costs have been dispelled. Operators who get the most bang for their IT buck place strategic focus on management, not money.

Duarte Begonha is a Principal in McKinseys Lisbon office. duarte_begonha@mckinsey.com

Giuliano Caldo is a Senior Expert in McKinseys Romeoffice. giuliano_caldo@mckinsey.com

Javier Garabal is a Principal in McKinseys Barcelona office. javier_garabal@mckinsey.com

Stphane Rey is a Principal in McKinseys Zurich office. stephane_rey@mckinsey.com

RECALL No 13 Leveraging Technology Sweating your assets: Riding the next wave of savings in the field force

13

02 Sweating your assets: Riding the next

wave of savings in the field force

Under pressure from ever fiercer competition and greater product complexity, telcos are seeking further technology potential to enable the next wave of field force productivity improvement. Leaders are finding that smart simulation across their entire operating model can reap swift gains without new investment. Many telecoms operators have implemented broad programs to enhance the end-to-end productivity of their field forces in the recent past. However, the battle for the wireline arena continues to rage, markets are maturing, and customers are demanding ever more complex products and networks. Telcos are faced with a dilemma. They urgently need to unlock further productivity to keep pace with customer and capital market expectations without increasing their budgets. A sophisticated technique involving lab-based simulation can unleash an additional 15 to 20 percentage points of further productivity from telco field forces without the burden of additional capex.

What went wrong? Many telcos added this new layer of technology without aligning the new systems to their core processes. The high pressure to deliver meant they often sacrificed functionality or focused on driving value in only one dimension. Insufficient investment in the alignment of systems and processes has left users either unable or unmotivated to tap the systems full savings impact. Some field organizations, frustrated by the perceived lack of value their GPS implementation offers, simply switch off their systems. Failure to provide targeted capability building and interlink the new technology with performance improvement and incentives has been an added stumbling block. However, driven by the need to mine further potential a few years on, telcos are now revisiting the sacrifices made during implementation and searching for ways to unleash additional value without significant additional investment. To do this, telcos must focus on the entire operating model (Exhibit 1), scrutinizing all their tools, data inputs, and core business processes and particularly the interactions between them.

A new mode of technology-enabled operations

Over the last five years, many major telcos have turned to technology enablement to improve the productivity of their network field force. In most cases, this has involved expensive up-front investment in tools such as automated dispatching systems, GPS, wireless handhelds and laptops, and other field tools. Such resource intensity has frequently led telecoms operators to neglect additional incremental investment in other areas ones that are crucial in maximizing the return on their technology investment.

Integrating operations around technology

McKinsey has tapped into a powerful technique for capturing this next wave of savings with its OpsTechLab. The key to keeping investment costs down and speeding time to impact by around 50 percent is to conduct IT-enabled simulations of client tools and process-tool interactions. One dimension is to constantly optimize the tools underlying algorithms to ensure greater input data quality. Another is fully aligning these tools

14

01

Field force performance can be optimized via improved integration of core operating model elements Field force operating model can be improved in many areas
Core operating processes Workforce management Work scheduling and tactical assignment of engineers to jobs Service delivery and management Closely managing job completion to ensure service delivery Resource/capacity management Managing staffing and demand levels to enable service delivery at lowest cost Performance management Measuring and managing performance at an individual and a group level

Tools

Scheduling algorithm Optimizing service and engineer utilization

GPS Pinpointing real-time locations

Forecasting model Estimating future demand

Engineer/technician tools Issuing handheld devices, laptops

Continuous improvement Identifying/addressing systematic hurdles to drive improved performance

Data

Engineer data

Customer data

Job data

Product data

Network data

SOURCE: McKinsey

with core operating processes. Integrating continuous improvement into these processes is also essential to ensure sustainable change. Priming system tools for new functionality. System tools are only effective when data inputs are reliable and relevant. Initial systems implementations often involve a thorough review of key inputs. However unless processes and resources are put in place to uphold data accuracy over time system tools gradually lose efficacy in line with the erosion of data reliability. Maximum automation is required for timeliness and efficiency. Two examples of high-impact improvement levers include daily technician job scheduling or dispatching algorithms and the automated infeed of updated technician skill profiles. Complex scheduling tools require thoughtful initial configuration and continuous attention to ensure their optimal performance. In practice, however, awareness of a scheduling tools true productivity drivers varies greatly. Changes made to accommodate business needs or specific regions also reduce a tools efficiency over time if its algorithms are not regularly re-optimized. Simulation in McKinseys OpsTechLab serves to test a scheduling engines performance, identifying areas for

improvement. Simulation tools use real data to create optimized scenarios that can be compared to the historical baseline. This approach allows telco leaders to tune up their dispatching tools by refining the algorithms they run on. Algorithms in the scheduling engine must be configured and weighted to reflect the true priority of assignments. This optimizes job allocation to meet service needs, reduces drive time, and maximizes technician utilization. Simulation bypasses the need to test potential improvements live prior to implementation. This lowers the risk of service failures, while speeding up implementation. Ensuring that scheduling tools receive an automated feed of skill profile updates is a second highly valuable but seldom used tweak to existing technology. This boosts the flexibility of resource deployment and maximizes scheduling algorithm effectiveness. It also increases the impact of training investments, as it ensures the organization deploys technicians to perform all the tasks for which they have been trained. OpsTechLab simulations often demonstrate a potential increase in jobs per FTE-day in the range of 8 to 10 percent, with travel time reductions averaging 20 percent. Integrating these smarter tools into core field force operating processes. Ensuring that data and tool

RECALL No 13 Leveraging Technology Sweating your assets: Riding the next wave of savings in the field force

15

reconciliation and updating are state of the art is only one side of the equation. Aligning processes to capture their full potential is the other. Again, up-front simulation of the impact is faster and far more cost-effective than field testing would be. Three examples of where the greatest benefits can be uncovered lie in the areas of workforce, service delivery, and technician performance management. Dynamic dispatch is the use of scheduling algorithms and GPS to drive workforce management. This enables dispatchers to assign field workers one job at a time throughout the day in the most efficient possible sequence, rather than the traditional method of loading them with a full days work up front. GPS gives the scheduling algorithm a more accurate picture of technicians real time locations compared with the jobs requiring completion. Scheduling algorithms then conduct multivariate optimization to deliver maximum technician utilization and lowest drive time, corresponding to service at the lowest cost a boon for both the company and the customer.

Compared with conventional full-day scheduling, using dynamic dispatch typically yields improvements in field productivity of 10 to 15 percent. This corresponds to the impact telecoms players are capturing as they implement real-world dynamic dispatch. And surprisingly most field technicians like it once they get used to it. A major source of stress for field technicians who are front-loaded or bulk-loaded is that one unexpectedly long job can upset their whole days schedule. Dynamic dispatch enables an organization to manage variability in task time completion using the entire field force, rather than asking a single technician to do this every day. Beyond providing a telecoms operator with cost savings opportunities, implementing dynamic dispatch also makes it possible to generate incremental revenues by enabling telcos to both offer and deliver differentiated service levels to their customers based on varying response times. This is one example of fully exploiting business processes to realize the return on systems and tools investment.

Unlocking impact: A case example

The experience of one North American telecoms player exemplifies the importance of optimization across the entire operating model, using an integrated approach. This operator already had a good workforce management process in place. They had deployed dynamic dispatch techniques and an automated scheduling algorithm. However, OpsTechLab simulations revealed that they were missing out on 5 to 10 percent in productivity savings by not regularly analyzing the effectiveness of their scheduling algorithm. By improving their algorithm configuration and allowing their technicians to perform all the jobs they were capable of, they could shorten driving distances between jobs and increase technicians value-added call time. The telco had access to all of this data from their tool, but did not have the simulation capability to interlink it and generate such savings opportunities. McKinsey helped update their algorithm using the findings from the simulation and also build the capability to regularly update and improve it. On the GPS front, managers were required to log into the graphical GPS system each day to ascertain the positions of their technicians and then crossreference with pages of data to spot any discrepancies. This system had two issues. First, it was hard for managers to determine whether technicians were really where they should be there were no visual displays to indicate irregular activity. Second, it required managers to sit at a computer tracking their staff in real time all day, which didnt happen in practice, given the many competing demands on managers time. A simple custom tool was created by the McKinsey team to highlight technician behavior inconsistencies for the managers attention at the end of each day, allowing managers to spend time where they should be coaching technicians. Optimizing the interlinks between tools and processes helped this telecoms player drive a productivity improvement of 10 percentage points while enhancing the return on their previous technology investment.

16

GPS-based scorecard
Scorecard with GPS data
Daily productivity Percent Tech 1 Tech 2 Manager
1

Task time attainment Percent 85 75 80

Productive time Percent 67 86 76

Total worked time Minutes/day 540 565 553

Late start Minutes/day 12 -10 1

Early finish Minutes/day 45 32 40

Time away from task Minutes/day 120 59 90

57 64 61

Definitions
Productivity Task time attainment Productive time Total worked time Late start Early finish Time away from task
1

Task time attainment x productive time, as a percentage (actual time to execute assigned tasks) / (standard task time for assigned tasks), as a percentage (Total worked time (late start + early finish + idle time)) / total worked time, as a percentage Average time worked during the given period, in minutes per day Average delay in departure time vs. expected start time, in minutes per day Average early return to overnight location vs. expected completion time, in minutes per day Average stationary time away from expected job location, in minutes per day

Managers metrics are the average of all assigned technicians

Another area where few organizations are capturing the full potential of tool/process alignment is the ability of GPS to provide real-time visibility on service delivery and service management. McKinsey has conducted several recent systems implementations with clients to define the requirements for installing real-time visibility of their service delivery. Close cooperation with vendors allows identification of the system requirements for smart delivery of such visibility without overwhelming managers and dispatchers. Real-time alerts, for example, notify dispatchers when service delivery is at risk, enabling them to contact technicians and offer support, or escalate issues to field managers. This gives dispatchers and the field a greater sense of shared responsibility in service delivery. Performance management is also greatly improved by expanding GPS functionality. Coaching and performance dialogs can be informed by richer data broken down into time on site, driving, and unproductive time, rather than by approximations and averages. Expected job times can be refined based on actual rather than estimated work times, providing standard task times or norms for specific types of tasks. Above is an example of a GPS-based scorecard that can be used by first-line

managers to identify drivers behind technician performance. It serves as the starting point for performance coaching with the technician and enables managers to customize their approaches. A manager could, for example, investigate high time away from task using retrospective location data. A manager might also recognize the need for additional technical coaching for technicians with high productive time but low task time. Enabling sustainable change with continuous improvement. Top telcos know that, with technology, standing still can mean falling behind. All elements of the operating model require continuous improvement to ensure maximum return on technology investments. Historically, improvements have been made reactively in isolation based on technical user group observations. The imperative now is to drive continuous improvement in an integrated way, using strong cross-functional teams, regular review and maintenance cycles, and an effective feedback loop from all stakeholders. Most telecoms players have technical specialists focused on a single aspect of the operating model: processes, tools, or data inputs. Telcos need to augment this domain-specific expertise with a small cross-functional

RECALL No 13 Leveraging Technology Sweating your assets: Riding the next wave of savings in the field force

17

leadership team that has a mandate of year-on-year improvement in overall systems performance. These systems optimization leaders must bring a specific set of capabilities to the organization. First, they need to have a leaders understanding of the business and its performance aspirations. They must have the problem solving capability to optimize both within and across domains using simulation tools, user group feedback, and regular system diagnostics. Finally, they need the program management skills to steer maintenance cycles and construct and orchestrate feedback loops. To start, these leaders need to develop regular maintenance and review cycles for their information systems algorithms, data input quality, and process-tool interactions with clear expectations for target system performance and the associated field productivity levels. Weekly, monthly, and quarterly maintenance schedules can be defined in detail, with clear responsibility assigned. Reviews can be incorporated into software release cycles to ensure quick responsiveness when additional development is required. In addition, vendor reviews and technology forums can be incorporated into an annual cycle to ensure the organization keeps abreast of external technology tips and trends.

Such leaders also need to build effective feedback loops from a broad cross-section of the user population. Realtime issue reporting from technicians, dispatchers, and management should accompany more formal user panels and lead user feedback. User feedback can serve as a source of ideas to be tried and tested using simulation. Regular feedback also ensures the organization is always learning ways to improve the effectiveness of its people through the intelligent adaptation of technology tools to better suit their needs.

The leaders driving this next wave of field force cost savings are doing this by squeezing the maximum out of assets they already have. Telcos are repeatedly finding they can quickly unlock 15 to 20 percentage points of additional productivity improvement from their existing technology investments. Simulation across multiple dimensions in a risk-free environment with a focus on better integrating data, tools, and processes is the fastest and most reliable way to deliver on the full promise of technology enablement.

Brant Carson is an Associate in McKinseys Toronto office. brant_carson@mckinsey.com

Kelli Fairbrother is an Associate Principal in McKinseys London office. kelli_fairbrother@mckinsey.com

Todd Scheidt is an Associate Principal in McKinseys Toronto office. todd_scheidt@mckinsey.com

RECALL No 13 Leveraging Technology Lean on the line: Improving ADM in telecoms

19

03 Lean on the line: Improving ADM

in telecoms

Once the domain of automotive manufacturing, lean is now transforming telecoms. Fine-tuning a telcos innovation engine can lead to big competitive gains in productivity, time to market, and quality. Given the current challenges of the telecoms industry (e.g., creating new services, improving quality of existing services), IT-enabled business innovation is becoming more and more important. However, the speed, cost efficiency, and quality of the innovation channel at many telcos leave significant room for improvement. Boosting performance in application development and maintenance (ADM) now tops the management agenda. McKinsey experience indicates that ADM performance improvements can have a significant impact. First, productivity gains in ADM of more than 30 percent are possible (equal to 1- to 3-percentage-point increases in company profit margins), leading to lower costs or greater capacity to deliver on business requirements. The impact of performance improvement, however, reaches beyond the ADM activity itself into the entire organization. As a result of their efforts, telcos have enjoyed time-to-market improvements of 10 to 25 percent for changes in price plans, bundles, and new development projects, and have seen reductions in defects in delivered software of 20 to 45 percent. These benefits apply to both IT departments as well as to separate business-to-business ICT services units that provide ADM services directly to customers. In the latter case, we have observed EBIT increases of around 5 to 7 percentage points for the ICT services unit.

A lean overview
While the manufacturing industry originally pioneered lean management concepts, other industries have followed and implemented these principles in their own business contexts. Lean management transformation is very practical as it invests 25 percent of the time on analysis and 75 percent of the time on implementation. The transformation is based on an iterative learning process where new ideas are tested, feedback is gathered, and improvements are implemented immediately. Over the past several years, application of lean management concepts to ADM has gained momentum. More and more companies are implementing lean management concepts in their ADM departments. Lean management focuses on five interlinked elements to improve overall performance, and these elements can be applied to ADM in much the same way as their traditional industrial application. The first two principles of lean management as they relate to ADM speak to the organizations systems. First, process efficiency is the way processes and resources are used and configured to create value while minimizing the cost to serve. Second, performance management comprises the processes, systems, and infrastructure needed to manage performance across ADM activities. The next principles relate directly to the human element of the organization. Mindsets and behavior describe the way the staff of the organization think, feel, and act both

20

individually and collectively. Organization and skills are about the extent to which processes are managed and supported by a skilled base of workers. At the heart of lean management the element that links all of its principles is the voice of the customer. This means the services that come from ADM activities are aligned to real customer desires. A lean management journey typically consists of three phases: Diagnostic phase (two to three weeks). This phase begins by generating an initial hypothesis regarding the main improvement levers. A fast scan of current working practices on the work floor (called a walkthrough) combined with high-level benchmarking serves as the basis. After that, a deep dive with lean management techniques to better understand root causes and solution directions begins. Future state design phase (two to three weeks). Here, several lean management solution archetypes (e.g., business requirements, iterative development, performance management, and testing efficiency) lead the organization toward a new way of working. The advantage of having lean management archetype solutions is a steep acceleration of implementation and impact. Implementation phase (around eight weeks). The new way of working is established and then refined over the given time period. The first visible elements that teams will encounter are whiteboards and daily briefing meetings. The daily briefing is a 15-minute, stand-up meeting that takes place in the morning, in which the team discusses the previous day, plans the coming day, and raises challenges/problems. During this phase, tactical implementation plans are also further developed. In general, about 10 percent of the expected benefits can be captured by the end of the first eight weeks of the implementation phase, which will continue for another nine to twelve months.

Diagnostic. The diagnostic began with a top-down performance review (benchmarking and walk-through technique) to generate an initial hypothesis on productivity, throughput time, and quality improvement. Subsequently, the diagnostic deep dives focused on validation of the initial top-down view, better understanding of root causes, solution directions, and achievable impact in the short to medium term. McKinsey used many different lean management diagnostic tools. A subset of the main ones used is included here. For the process area, the focus lay on time spent on value-added versus non-core activities (overall process efficiency analysis), reduction of work inflow (valueadded work analysis), and process mapping to discover rework loops and waiting time (value stream mapping analysis). For the performance management area, emphasis was placed on assessing KPIs, visual aids (within lean management also referred to as Kanban), and review cycles. The organization angle looked at gaps between employee skills required and those available. To diagnose mindsets and behavior and the voice of the customer, employee surveys helped evaluate how staff acts and feels, whereas a process maturity assessment focused on the quality of the customer interactions and on talent management. The diagnostic revealed that project and support efficiency could be increased by 22 to 45 percent based on the actions listed in Exhibit 1. The diagnostic also indicated potential to lower the share of projects with budget overruns from 45 percent to between 20 and 35 percent, while cutting the cost price of resources by around 15 percent to bring it in line with market levels. Finally, the diagnostic identified potential to double the current ratio of people working offshore versus onshore and cut throughput time by about 15 percent. As a realistic goal, the organization committed itself to reduce costs by 25 percent over 18 months. Future-state design. Based on the diagnostic, this phase focused on four large design archetypes: business requirements, testing, management and support, and performance management. The business requirements archetype was employed to reduce rework, prevent budget overruns due to incorrect scoping, and drive down non-productive development hours of analysts, while enhancing the quality of handover when work moves offshore. This design archetype focused on the full process chain from customer

Lean ADM in telecoms: A case study

One telco with whom McKinsey has worked has an ADM department with more than 1,500 FTEs. The organization is now well on track to reaching (and even exceeding) its improvement targets. The following recounts their lean ADM journey.

RECALL No 13 Leveraging Technology Lean on the line: Improving ADM in telecoms

21

01

Efficiency improvement potential ranges from 22% for average to 45% for top performers Efficiency improvement potential ranges from 22% for average to 45% for top performers
Efficiency improvement initiatives for average and top performers Reduce time focused on non-core production activities Prevent rework resulting from unclear and changing business requirements Lower support staff efforts like administration, finance, and HR Increase testing efficiency Ensure sufficient scope of control for project management Raise utilization by ensuring staff is working on projects Total efficiency improvement Waste as a percentage of total time 9 - 15 5 3-6 3-6 2-6 0-7 22 - 45

SOURCE: McKinsey

insight research and collection to change management in requirements after initial sign-off. Here, use cases were selected as the method for business requirement design. The main rationale for choosing use cases was that, when applied correctly, this forces completeness and a proper level of detail and it is easily understandable for business. It also immediately feeds into the performance management initiative through use case points a metric to measure productivity output. Testing to reduce non-productive hours and enhance overall efficiency was the second archetype that the telco used. This design phase focused on improving the interaction between development and testing activities through tighter quality control of unit testing output. It also concentrated on shifting large portions of system testing to earlier points in the development cycle (instead of a big-bang testing approach at the end of the development cycle) to enhance feedback to the developers. Another important focus of this phase emphasized improvements to the testing activity itself based on automation of integration/regression testing for parts of the system code that remained relatively untouched over the different releases. In further improving riskbased testing, the telco ultimately will decide on the amount of testing effort to invest based on the codes functional and technical risk profile.

The management and support archetype was employed in order to achieve a reduction in the number of support staff and project managers. The main reason for the large numbers of support staff was the abundance of manual finance and administration tasks. To drive efficiency, one objective of the design was to implement a project management automation tool and consolidate the remaining support work. For project managers, the design focused on increasing efficiency in communication interfaces with other parties within the organization (primarily participating in meetings and preparing reports). Finally, the performance management archetype was used to buttress the lasting impact of all initiatives. On a project management level, the design consisted of a large whiteboard that showed planned versus realized work on a person-by-person basis (visual work flow board) as well as a burn-down chart that visually tracked overall team progress over time. Each morning, team members would discuss outcomes of the prior day and plan for the day ahead. On a weekly and monthly basis, a richer set of KPIs showed progress on other dimensions such as utilization, fully-loaded cost per FTE, defect density, effort overruns, and span of control. On a management level, a series of review meetings across the organization was set up to discuss the KPI reports and served as a forum for management to decide on actions based on facts.

22

A very conscious decision was reached to make the detail design of these archetypes line-led, meaning that line management drives the design choices described above supported by lean management experts. This ensures better organizational acceptance for the solution and a smoother subsequent implementation. Implementation. The telco implemented its lean process in three distinct stages. The first implementation stage was to test the design. To ensure the perfect fit, McKinsey worked with the organization to implement the full future-state design in a project launched recently. After some marginal fine-tuning to accommodate organization-specific requirements, the solution was ready for full rollout. The second stage of implementation was to develop the organizations capabilities. The telco started its capability-building effort based on adult learning doctrines under the guiding principle that it is important for people to first understand the reason for change before training them in new skills. The business requirements and performance management initiatives required the greatest training effort given the use of new tools like use cases, use case points, burn-down rates, and visual work flow boards. To develop the necessary capabilities, the telco conducted workshops for analysts and project managers to build awareness of the gap in the quality of current versus best practices this, for them to internalize the importance of change. The telco also offered e-learning and selected expert classroom sessions to educate employees on the new way of working as well as to develop internal coaches who could continue the training efforts beyond the individual sessions.

The third and final implementation stage was to rapidly scale up the solution. After testing the design, McKinsey worked with the telco to finalize the blueprint design material. It is important that this blueprint consists of practical material that people will use daily (like standard operating procedures and quick reference guides) rather than lengthy manuals that are hardly ever used. In only a couple of waves, the capability-building program covered the entire organization. At the end of each wave, everybody immediately adopted the new way of working. Rolling out the new approach rapidly ensured that the full organization works uniformly. The results generated so far have boosted confidence throughout the organization that the lean ADM journey is indeed a very rewarding one. It also remains very high on the top management agenda and visibly so. All of these elements are important in ensuring that impact is sustainable and the organization steadily moves toward continuous improvement, even after the initial effort. For this telco, the lean ADM journey has just begun.

Telcos can obtain faster and cheaper innovation capabilities with better output, leading to a competitive edge. Applying lean management principles to their ADM activities is a tried and true means to this end. Lean management focuses on process and the associated performance management optimization, but goes even further to develop capabilities, alter mindsets, and ensure continuous improvement.

Joris Hppener is an Associate Principal in McKinseys Tokyo office. joris_hoppener@mckinsey.com

Stphane Rey is a Principal in McKinseys Zurich office. stephane_rey@mckinsey.com

Kevin Wei Wang is a Principal in McKinseys Shanghai office. kevin_wei_wang@mckinsey.com

RECALL No 13 Leveraging Technology Fixing break-fix: The power of lean ICT transformation

25

04 Fixing break-fix: The power of lean

ICT transformation

Operators focused on the business-to-business ICT space can benefit significantly from lean transformation of their field force operations. In recent years, the field force has become a critical competitive element in telecoms players high-volume, standardized customer service operations. Operators have learned to leverage this element to also drive value based on cost efficiency and quality. But a very different situation persists today for most players serving the large B2B information and communication technologies (ICT) customer segment. The service delivery chain in B2B ICT differs from B2C telco service operations in several respects. First, service level agreements (SLAs) for large accounts usually encourage customization. This means that attempts to standardize the process are often thwarted from the outset. Another factor is that service delivery windows are much shorter: most repair tickets involve turnaround within 8 or 24 hours. Beyond this, the work requires a higher share of specialized skills. The B2B ICT segments broad hardware portfolio also leads to greater logistics complexity, which the field force has to support alongside same-day service provision. For all these reasons, lean transformation has traditionally been considered less applicable. This belief, however, is a fallacy. To demonstrate the power such a transformation can have, this article presents a sanitized case study. It involves a leading integrated EU telecoms/ICT player with an enterprise business unit that also installs and maintains the ICT

systems of large B2B customers. The company aimed to improve profitability and customer satisfaction in its ICT hardware repair operations by boosting the productivity and quality of its service operations. To accomplish this, company managers decided to adopt an end-to-end (E2E) focus on the delivery chain for repair activities conducted at the customer site. The results demonstrate that lean transformation in this area can raise productivity by some 40 percent and at the same time realize quality improvements of around 10 percent.

Working end-to-end to maximize improvementpotential

The lean field force transformation focused on five areas (Exhibit 1): contact center, service qualification, planning, field engineers, and logistics. These areas constitute the core E2E process for on-site ICT repairs (mainly hardware). The company exhibited all the typical differentiating characteristics outlined earlier. About 40 percent of the tickets were for same-day repair (a maximum service window of 8 hours); another 40 percent were for resolution within 24 hours. Despite the challenges, the company achieved significant impact during an eight-week pilot phase. The results were measured using a detailed dashboard, which also showed that these results were being sustained over the long term as well. Field engineer productivity, for example, increased by nearly 40 percent and quality measured in terms of the share of repair tickets fixed right the first time rose by over 11 percent. Planning effectiveness the number of tickets per

26

01

The end-to-end (E2E) on-site repair process addresses five key areas The end-to-end (E2E) on-site repair process addresses five key areas
E2E hardware ICT repair service process Customer interaction Customer reports incident Telephone, e-mail, EDI, portal) 1 Open ticket Complete customer serial number, etc. Contact center 6 FTEs Diagnostic contact customer or send technician Contact customer to make appointment if requested Execute repair at customer site

Internal process (~ 75 FTEs) Activity

3 Completeness check Technical diagnostic Contract check Vendor warranty 6 FTEs SLA check Matching ticket with technician and parts Delivery speed mgmt. Planning 6 FTEs

4 Travel Parts pickup On-site execution Return of parts 37 FTEs

Function

Service qualification

Field engineers

Parts advice 5 Procurement Stock management Transport coordination Warranty claims

Parts confirmation

Depot/courier (handshake)

Logistics
SOURCE: McKinsey

18 FTEs

engineer per week climbed by nearly 35 percent and overall productivity went up by roughly 36 percent (suggesting very similar latitude for FTE reduction).

Getting started: A two-week lean diagnostic

Field force transformation begins with a diagnostic phase. An effective diagnostic surfaces a wide variety of issues that the transformation teams need to work through in all these areas. Key diagnostic tools used include value stream mapping, field observation, and variability and disruptions assessment (also across peers). The main findings in each area were revealing. Contact center/service qualification. The company discovered that the customer information collected was often either incorrect or incomplete (e.g., end-user workstation identification and customer serial numbers) and that agents were not always making direct contact with end users. More effective remote diagnostic scripts and tools were needed. The contact center and qualification also lacked a clear process for performance reporting, such as the number of clean tickets passed on to planning and field engineers. Planning. Direct observation showed: engineers spent less than 20 percent of a typical day on value-added ticket time (Exhibit 2). Around one-fourth of their time

was dedicated to administrative tasks. The main downtime in the process was for travel, administrative activities, waiting, and searching for the client manager or exact workstation location on the customers premises. Field engineers. The diagnostic spotlighted issues in both field engineer support and peer performance. Field engineers faced irregularities in over 80 percent of total tickets received. These mainly related to logistics (over 60 percent) and planning (around 10 percent). On average, this already represented 35 minutes of waste per ticket, equal to around 15 percent of the organizations total daily field engineer capacity. The quality of this support definitely required significant improvement effort. Where productivity was concerned, a peer comparison diagnostic revealed a large spread. The top 25 percent were completing roughly twice as many tickets per day as the lowest quartile. One of the drivers was admittedly the geographic concentration of tickets within a region. Many low performers were in more sparsely populated regions, resulting in increased travel downtime. But even within the top quartile, the spread between the top and bottom 25 percent amounted to 2.4 successful tickets per day. High variability in ticket completion times, even when the incidents reported were very similar, suggested significant optimization potential.

RECALL No 13 Leveraging Technology Fixing break-fix: The power of lean ICT transformation

27

02

Value-added time accounted for less than 20% of an engineers day Value-added time accounted for less than 20% of an engineers day
Time spent on 1 day, based on observations during go and see field visit Number of hours (rounded) Non-value-added activities before/after tickets 8.0 1.0 1.0 1.5 4.5 1.0 1.0 0.5 Lunch 0.5 Waste in process 1.5 Valueadded ticket time -81% Non-value-added activities during ticket execution

Work-home Warehouse Nontravel admin flexible planning Description From home to warehouse Find and check parts for the day Check and correct ticket for the route day Tickets finished at 2:00 p.m., but no additional work possible

Gross time Travel time Admin for tickets From Closing local calls ware Logging house to data1 customer location Between customer locations

Lunch at Search- Diagnosnoning for/ tic client calling Repair location customer Testing Phone calls during ticket execution

1 Assigning ticket and allocating tasks to specific minutes SOURCE: McKinsey

Logistics. Logistics encompasses procurement, stock management, transport coordination, and warranty claims. Around 60 percent of corrective actions in logistics were courier-related (such as the courier arriving too late), with couriers accounting for about 25 percent of total logistics costs. Waste from frequent logistics disruptions included parts not delivered to the right local inventory locations or not assigned to a ticket (PDA synchronization problems, etc.). Some 5 to 8 percent of parts were defective on arrival. Around 50 percent of parts shipped to engineers on site are returned untouched, a sure sign of overqualification. Another area requiring optimization was stock turn: analysis showed that over 90 percent of stock-keeping units were only used twice a year at the most.

engineers, and logistics), as well as ideas to streamline the E2E nature of the work flow. Boosting contact center/qualification quality. At the contact center, one initiative was to review and optimize the scripts used for capturing critical customer data, such as the end users mobile number, customer serial number, location, and the key contact. In the qualification phase, initiatives included reducing follow-up tickets by enhancing attention to detail and keeping staff up to date on newly introduced contracts and models. Optimizing planning and capacity management. Numerous proposals were generated to improve field force planning and capacity management. Every detail of next-day tickets was set up the day before. The planning team mapped all tickets and to dos on a planning board that they checked regularly at noon and at the end of the workday. Unproductive time was balanced out with much more precision: engineers not fully scheduled for the following day were assigned to a flexipool for either part of the morning or afternoon. One productive use of their time is to inspect repair parts at the warehouse to check quality and to give suppliers feedback. Initiatives for field engineers included introducing a normed ticket time based on performance management to reduce variability in ticket completion times. This encouraged an effective decrease in average

Powering up for higher performance

The company used the detailed insights of the diagnostic phase to develop improvement initiatives based on the lean philosophy along four productivity levers: capacity management, planning management, coaching for results and performance, and reducing disruptions within the E2E service delivery chain (Exhibit 3). Ultimately, the team identified more than 40 initiatives during the two-week diagnostic, spread across all five areas (contact center, qualification, planning, field

28

03

Several specific initiatives resulted in significant impact Several specific initiatives resulted in significant impact
4 productivity levers Capacity management Balance number of available field engineers with expected ticket volume Examples of initiatives Minimum number of FTEs in fixed pool of engineers Flexi-pool to buffer peak load Engineers shifted from FTE pool to flexi-pool Next-day tickets planned at end of day Planning team checks scheduling board every day at noon and end of day Norm time reduction Weekly performance review between field manager and team Field manager coaches worst performer individually at least once per week At least 1 contact with end user during E2E ticket delivery Planning and logistics team sit together in openspace setup ILLUSTRATIVE

Planning management Increase field force effectiveness

Fill each engineers free time with productive tasks Reduce travel time Reduce variability in engineer performance via coaching and performance management

Coaching for results

Reduce disruptions within chain

Reduce disruptions in ticket execution Reduce follow-up tickets

SOURCE: McKinsey

ticket time. With 50 minutes estimated as the weighted average time achievable, the team determined it could generate a 25 percent reduction in average ticket time. An alternative planning and dispatching method is based on a dynamic approach. Here, field engineers are assigned one job at a time throughout the course of the day. This method makes it possible to manage variability in ticket completion time across the entire field force. However, when highly specialized hardware is involved that has to be taken to the site by the field engineers, planning a day ahead is required to distribute the appropriate parts to the field engineers. Streamlining logistics. Logistics initiatives focused on aspects such as improving the coordination of parts sourcing and transportation by courier, in alignment with field engineer planning. The parts drop was organized at a single location and bundled into a single role. Previously, this had been organized across different teams at separate physical locations. Other ideas generated were to invest engineer time into ensuring that parts receiving at local warehouses was optimally orchestrated. A 6:00 a.m. check was conducted on parts that had been delivered overnight and central pickup was organized. Prior to this streamlining, field engineers would pick up parts individually at non-assigned warehouse locations.

Nurturing lean thinking for sustainable change

Beyond optimizing processes, aligning staff mindsets and behavior is just as important, anchoring the concepts and philosophy of lean continuous improvement from an E2E angle. Opportunities for manager role modeling were set up alongside a constructive coaching environment in which staff could refine their skills and capabilities. This skill building dovetailed with performance management, helping drive long-term sustainability. One initiative was to introduce an integrated E2E dashboard that tracked progress on a weekly basis together with weekly meetings that addressed any E2E issues. Other actions focused on improving team interaction. A 15-minute get-together at the beginning of each shift provided a daily update on team productivity, quality, and morale, while offering a forum to handle any outstanding issues and gather new team ideas on performance improvement. At the end of each week, half-hour team discussions were held to maintain the momentum, while benchmarking quality and productivity performance against past results. The companys successful lean field force transformation taught a number of important lessons on how to best tap into the innate but often latent power of the

RECALL No 13 Leveraging Technology Fixing break-fix: The power of lean ICT transformation

29

field force organization. Where scope and setup are concerned, it is essential to work end-to-end to avoid misalignment on KPIs and SLAs at the subunit level. In execution, identifying the right drivers and breakpoints is crucial. Linking thresholds and targets to real financial impact as well as customer excellence is also a strong enabler. And ultimately, imprinting lean principles on the companys DNA by refining employees soft skills is perhaps the greatest differentiator: this is the key to a continuous improvement culture.

Bart Delmulle is an Associate Principal in McKinseys Brussels office. bart_delmulle@mckinsey.com

Leon de Looff is a Principal in McKinseys Amsterdamoffice. leon_de_looff@mckinsey.com

Peter Verboven is an Associate Principal in McKinseys Brussels office. peter_verboven@mckinsey.com

Matthias Winter is a Director in McKinseys Zurich office. matthias_winter@mckinsey.com

RECALL No 13 Leveraging Technology Outside in: Leveraging outsourcing and offshoring to transform IT

31

05 Outside in: Leveraging outsourcing and

offshoring to transform IT

Many operators outsource and offshore their IT, but few capture the value that a full-scale transformation can unleash. How to become one of the few. Conventional IT outsourcing and offshoring (O&O) can deliver significant medium-term improvements, but benefits ultimately flatten out because companies retain the same legacy IT platforms. Implications are severe, including longer time to market and growing operational risk. Industry players have begun to address this issue using innovative O&O models where vendors drive, co-invest in, and accelerate modernization in the process realizing incremental cost savings of around 20 percent. These models require a different O&O mindset that builds on shared objectives, employs joint telco/vendor implementation roadmaps, and engages in common value tracking. First, however, it is important to see where conventional O&O setups often go awry.

little in the way of real IT transformation has transpired. While these players may enjoy highly competitive unit prices and good utilization levels, they continue to operate in their accustomed environment: their delivery models, their systems, their data models, and their technical architectures remain pretty much unchanged. Companies at this stage of O&O maturity often find that burdensome legacy has begun to consume the benefits of their O&O arrangements. Expenditure on their IT operations keeps rising because each dollar of new functionality adds roughly 10 to 15 cents in additional annual maintenance costs. This means they have less in their wallets to develop innovative, business-enabling IT solutions given the resulting cash flow constraints. They also have to contend with an increasingly complex time- and money-consuming design and testing process which typically leads to slower time to market and less project delivery predictability. These serious repercussions can ultimately jeopardize an operators competitive position. However, while the need for additional investment in IT transformation is clear, circumstances can make it impossible for CIOs to release the cash needed. First, the business case even for evolutionary IT transformation is relatively challenging since the payback horizon may extend beyond the CEOs mandate period. Second, funds for capex are still scarce in the current business environment, and customer-facing projects will inevitably receive the capex green light first anyway. And third, vendors in most existing O&O contracts often find it advantageous to maintain the status quo rather than contributing to removing complexity.

The downside of conventional O&O

Well-structured and well-negotiated arrangements can yield cost savings of some 20 to 30 percent over three to five years. Primarily, transferring activities offshore, capitalizing on vendor process improvements, accessing larger resource pools, and upping infrastructure asset utilization serve to achieve this. Generally, contracts also include well-designed demand management mechanisms for telcos to capitalize on stricter, more professional project and service level prioritization. However, when one examines these deals two to three years after the ink on the contract has dried, quite often

32

01

A service architecture view helps translate objectives into the right transformation path A service architecture view helps translate objectives into the right transformation path
Service architecture Business processing Application development Value chain Application maintenance Application operations Infrastructure operations Enterprise resource planning Sourcing strategies (examples) 1 2 1 2 3 4 5 5 3 6 4 5 6 7 Billing Group functions Single-vendor co-financed modernization Single-vendor data center/ servers Single-vendor end-user computing services Preferred offshore vendors Two prequalified niche providers Software as a service Business process outsourcing

Business domains (examples)

SOURCE: McKinsey

The power of O&O to drive and accelerate ITtransformation

A multi-lever transformation is required to crack the curve described above and move to the next level of IT performance. Telecoms players need to move from ad hoc processes with vague contours to well-defined services, using configurable standard packages rather than customized (largely unsupported) software. This will result in a gradual shift from fragmented data models, tailor-made application programming interfaces, and logic-heavy middleware to modular, servicesbased architecture. The landscape morphs from dozens of legacy technology platforms to a handful of standard technology stacks. To achieve this type of true IT transformation without significantly increasing their IT spending, operators have to take a much deeper and more collaborative approach to O&O. Instead of merely pursuing tactical business objectives like cost, flexibility, or quality, participants need to share and agree on business priorities regarding the future evolution of each IT domain. To capture the maximum amount of value, managers should step beyond attempts to squeeze margins. Instead, they should ensure common understanding of the detail the new platform will entail, with exact itemization of the implementation plan, KPIs, and

resources, for example. This allows them to fully gauge the impact of individual transformation initiatives. Conventional O&O tactics regarding the deal itself tend to be static, such as one five-year deal for all application maintenance: they do not reflect the varying priorities required in specific IT domains. Instead, telcos need to tailor their sourcing strategy to fit domain-specific priorities (multi- versus single-sourcing, for instance, and transformation versus pure offshoring), creating the right incentives both internally and externally.

How to set up a joint operator/vendor model

The first step is to develop an internal perspective, identifying where pursuit of IT transformation via O&O would make sense. Telcos need to recognize that objectives such as cost efficiency, competitive price development, or fast time-to-market performance will translate into different transformation paths depending on which part of their IT landscape is in focus. Taking a service architecture perspective will help teams analyze this issue in a highly structured manner (Exhibit 1 ). A comprehensive sourcing strategy then needs to be developed that aligns with these objectives. The foundation should be a credible economic model that assesses modernization priorities per domain, establishes a baseline, estimates the financial impact, and designs

RECALL No 13 Leveraging Technology Outside in: Leveraging outsourcing and offshoring to transform IT

33

02

Cost/price development should decrease to below the baseline within a relatively brief period if carefully architectured Cost/price development should decrease to below the baseline within a relatively brief period if carefully architectured
Cost/price development Percent of baseline 1 1 Build your internal view of the transformation roadmap and economics Negotiate joint roadmap with vendors, maximizing offshoring, operational excellence, and scale levers Agree on deal duration and financing to trade off up-front capex against long-term opex

2 100 2 3

Years

SOURCE: McKinsey

KPIs to quantify the effects. A telco will aim to concentrate volume with fewer providers, with niche vendors managed by the prime vendor in the domain. The heart of the strategy needs to be maintenance contracts with shared transformation: long-term maintenance is tendered to vendors, with the vendor co-investing in a shared modernization plan that locks in the transformation objectives. This will involve negotiating a joint roadmap with vendors that maximizes offshoring, operational excellence, and scale levers. It is vital to provide incentives for vendors to remove complexity. The provider may, for example, take over legacy systems from the clients technology platform. In this situation, the client should in addition to negotiating prices for specific services also create incentives for reducing the cost base. Awarding a long-term commitment that involves the maintenance of systems, once installed (rather than just a delineated project), is often a major incentive in itself. The clear scope and in-depth involvement of participating in an operational transformation in a strong relationship also gives the vendor greater assurance of capturing profitability over time. A further crucial negotiating point is the deal duration and financing to trade off up-front capex against longterm opex, making use of all value drivers (Exhibit 2 ). The running costs of a simpler, harmonized system are

Radically reducing complexity without increasing spend: An O&O case example

One operator collaborated with preselected vendors to first create domain-specific evolutionary architecture roadmaps. Then, they asked the vendors to design a model that maximized their ability to execute their roadmaps at minimum cost. Vendors competed on multiple dimensions: price, quality, credibility, and timing. The winning plan was contracted together with a long-term application maintenance agreement so that a significant share of the transformation was funded by the vendor. This operator is on track to see significant IT savings of well over 20 percent by fully leveraging offshoring and lean savings that go beyond the 30 percent already achieved from the previous O&O setup. Since the transformation modernizes the application architecture, the company can shut down several expensive legacies, beginning its move from proprietary vendor stacks to open platforms. The knowledge offshore vendors gain from maintenance and transformation work will also ultimately lead to a far better offshore application development ratio, cutting time spent on design and testing.

34

of course far lower than those of a legacy platform later down the line, which benefits a provider in a long-term contract just as much as the operator. Signing the contract with the vendor is just the beginning of the journey: the deal structure and governance of this type of multi-vendor landscape need careful planning to lock value into the application maintenance setup. The roles assigned to prime vendors, niche vendors, and operations require overall supervision by a coordinator acting as a single point of contact, with a clearly defined role. The added challenge is that this is not one overall transformation so much as the execution of many intricate mini-transformations, and requires leading-edge implementation. It is important, however, that the operator always retains the strategic and archi-

tectural prerogative, setting and monitoring the direction of the transformation. Pursuing a deeper, more collaborative approach to O&O can give operators the opportunity to achieve a fundamental transformation of their IT environment. Their O&O involvement should always be seen as part of an overall business transformation, led ideally by their business rather than their IT community. Operators are outsourcing ever more of their operations as they progress up the value chain. Transformational O&O enables and accelerates this process, but it also moves providers work up the value chain, creating a win-win situation for both operator and vendor.

Andr Christensen is a Principal in McKinseys Toronto office. andre_christensen@mckinsey.com

Martin Lundqvist is an Associate Principal in McKinseys Stockholm office. martin_lundqvist@mckinsey.com

Tor Jakob Ramsy is a Director in McKinseys Oslo office. tor_jakob_ramsoy@mckinsey.com

RECALL No 13 Leveraging Technology Fast forward to success: Managing next-generation IT architecture

37

06 Fast forward to success: Managing

next-generation IT architecture

Common pitfalls in telcos IT architecture management are all too often a serious roadblock to superior business performance. Operators can bypass these by applying learnings from their most successful peers. Telecoms operators are among the highest IT spenders across all industries, relying heavily on IT to enable their existing business as well as to open up new revenue streams. However, the impact of IT-related factors on a telcos business performance often remains unquantified in the breakneck race to stay ahead. McKinseys annual telecoms IT benchmarking survey provides strong empirical evidence of how dramatically telcos IT architecture impacts their business performance (Exhibit 1). Significant performance gaps are apparent between operators with lean IT architecture and those with a highly complex and fragmented landscape. The difference in time to market, for example, is striking: two weeks versus 32 weeks for a similar product bundle. IT spend, too, is around 70 percent higher for laggards than for best-practice mobile operators. The survey revealed three key differentiators in top performer approaches to their IT architecture: A consolidated application landscape per domain. The number of applications per domain clearly appears to influence an operators ability to drive the efficiency and effectiveness of their operations. The survey showed that operators with fewer applications perform substantially better than comparable competitors with a large number of applications. Best-performing mobile telcos

have 97 applications per domain on average, while laggards have 168. Use of standard software packages. Standard package functionality enhances business agility and drives down operational cost. The IT architecture of best performers is predominantly based on standard software packages in key domains, whether billing, CRM, or ERP. Mobile best performers spend 45 percent of their total software budget on software packages, while the figure for laggards is only 20 percent. These players rely on proprietary solutions for the rest. High level of data integration. Highly integrated architecture greatly enhances a telcos capability to drive product innovation and successfully commercialize their new products. The higher the degree of data integration (fewer customer databases, for instance), the shorter the operators time to market. This translates into a greater ability to push market innovation and gain competitive advantage. The large gap in these factors is not just found between attackers and incumbents, but also within attacker and incumbent groups. This evidence indicates that keeping IT architecture lean, leveraging standard software solutions from leading vendors, and rigorously managing data across the enterprise all seem to pay off in terms of shorter time to market, higher flexibility, and lower costs on both the IT and the business side. Many telcos fully recognize the downside of having a complex, fragmented IT architecture, but are trapped in

38

01

Strong evidence shows that IT architecture contributes to superior business performance Strong evidence shows that IT architecture contributes to superior business performance

Best-practice telcos Number of applications (mobile) per domain Number of client databases 97 2-4

Telcos lagging behind 168 > 10

Time to implement a similar product bundle IT spend as percent of revenues (mobile)

2 weeks 4.4%

32 weeks 7.4%

SOURCE: McKinsey

the vortex of spiraling complexity. The speed of change leaves them no time to integrate their architecture as they launch new multimedia offers and fixed mobile convergence products, use multi-channel delivery, refine their CLM capabilities, and enable real-time provisioning. Under such dynamic market conditions, keeping their IT architecture under control and avoiding unbridled complexity growth is an increasingly challenging task. The sections that follow outline the typical hazards of IT architecture management as well as effective tools and mechanisms for its sustainable optimization to create value throughout the business.

architecture. The result is numerous sets of applications with duplicated functionality, siloed and fragmented databases, and custom point-to-point interfaces. This not only applies to incumbents that have grown a complex legacy landscape over many decades; such practices are equally widespread among mobile attackers. Weak link between business and IT. While it is widely recognized that the business should take an active role in driving IT architecture evolution, CIOs and IT architects normally manage their IT architecture with little business involvement at the single application and project level. This means only limited levers are available to systematically transform the IT architecture to achieve sustainable business benefits across the enterprise. Mega-project approach. The decision to engage in multiyear transformational mega projects may be influenced by the promises of large vendors and integrators to manage implementation complexity. It is naive, however, to believe that business requirements will remain stable over the lifetime of a multi-year transformation. And statistics prove this, highlighting the low success rate of mega projects. The step-by-step approach of transforming domain by domain is much more resilient to changes in business needs and tends to deliver sounder benefits. Using a service-oriented approach can serve as

Typical pitfalls of IT architecture management

The IT benchmarking survey mentioned earlier as well as McKinseys experience in numerous projects with clients have highlighted a number of common pitfalls in telcos management of their IT architecture. All of these result in roadblocks obstructing business performance. Short-term perspective. Most telecoms operators IT architecture evolves driven by short-term business requirements in specific domains, with little regard for the long-term impact of this on the companys overall IT

RECALL No 13 Leveraging Technology Fast forward to success: Managing next-generation IT architecture

39

02

McKinseys telecoms domain map synthesizes a telcos functional capabilities McKinseys telecoms domain map synthesizes a telcos functional capabilities
Customer interaction Customer management Channel management Customer service Sales Marketing Billing Wholesale/service providers Wholesale/service provider management Wholesale/service provider performance management Wholesale/service provider settlements and payments Customer-facing
SOURCE: McKinsey

Management of product lifecycle, services, and resources Product lifecycle management Strategy and planning Capability delivery Service management Service brokering/orchestration Service quality management Problem management Resource management Provisioning and delivery Mediation and reporting Data collection and distribution Trouble management Performance management Resource enablers Service enablers Session control Configuration management Development and retirement

Support Enterprise support Supply management Risk management Stakeholder and external management Finance/controlling Effective enterprise management HR Strategic enterprise planning

the key component to create a common basis and allow the independent replacement of components or entire domains without affecting other domains. Deprioritized IT integration. A telcos IT is constantly under pressure to deliver new functionality, enabling new products, channels, and processes. The integration of existing applications is often lower down on the agenda, resulting in loss of efficiency, in errors, and in extended time to market. How can telcos take the fast track to successful architecture management, avoiding these snares? Any advice should cover incremental changes as well as major transformations, since enterprise architects have to cope with both of these factors at the same time. Numerous small improvement projects require guidance on architecture compliance, while the fewer but much larger transformation projects need full attention to develop the right concepts and lay the foundations for more radical change.

management of their IT architecture to create maximum value. Maintaining the leanest possible architecture via streamlining is one. Another is the constant prioritization of business requirements. Focusing on enterprise-wide integration is important, too. And an appropriate organization is just as crucial, with the right management team in place and compelling incentives for compliance. Streamline to ensure the least possible complexity and greatest value. A number of elements are indispensable in achieving greater architectural simplicity and value. The cornerstone is a well-defined business strategy. This should drive all other IT-related decisions and will also embed the longer-term perspective that is so often missing. Is the telco going to move into a fiber rollout, for example, and if so when? If key elements of the strategy are undefined, its IT architecture cannot be aligned to provide the highest value. A domain-based approach is essential. The domain map (Exhibit 2) serves as a single structure into which all solution elements fit, providing a nomenclature for business capabilities with the fewest possible interfaces between domains. It can be used to map both applications and data. While it is sensible to have a high-level view of the target end state, more detailed planning

Key factors for developing sustainable, value-driven IT architecture

The IT practices of best-performing telecoms operators reveal a number of common themes underlying the

40

Kick-starting the IT architecture journey to maximize value

Gaining insight into the key factors that contribute to success is one part of the equation. The other is to assemble a team able to apply them effectively. A European mobile telecoms operator recognized its chance to reap significant business benefits by leveraging IT architecture principles on its transformation journey. The telco also understood that people matter most. Heavy competitive pressure and unwieldy legacy systems were forcing this operator to renovate their customer domain and integrate it more tightly with their other customer-facing functions. They created an architecture group to oversee the project, and an experienced leader was selected for this function with expertise as well as excellent connections on both the business and IT sides. The team he assembled had in-depth experience in developing conceptual functional architecture models that could be realized using standard software components. Top management empowered the group to attain transparency on all aspects of implementation, from identifying business requirements at the outset to rolling out the solution. This included gathering information on all ongoing activities in the customer domain to spot interdependencies and inconsistencies between initiatives. While the local project management office had overall responsibility for the program, the architecture group was involved in all major decisions and had extensive rights to veto problematic developments, such as over-customizing the campaign management solution or taking shortcuts by connecting shared systems using point-to-point interfaces. The architecture group that evolved from the project structure has now been implemented as a permanent enterprise architecture management unit functioning between the business side and IT delivery. The group has developed an overall target blueprint that helps guide all current and future requirements. They also oversee the existing application landscape to ensure that any change made to the architecture complies with the target blueprint. Although the journey is still just beginning, the benefits of a well-designed, integrated architecture are already evident. They have created full transparency on application redundancy levels in the different functional domains and made concrete plans to decommission systems, replacing others with standardized software packages. As a result, IT operating costs are expected to decrease, and more importantly development capacity will be freed up for business innovation. The consolidation of customer data shows first signs of unleashing measurable business value: the growing transparency enables improved up- and cross-selling, as well as greater customer satisfaction.

should be broken down by domain. This reduces planning complexity and gives domain architects the freedom to advance their landscape at their own pace. Another practice preferred by top performers, as mentioned above, is to use standard software within each of the domains just outlined to cover most business functions in the architecture transformation. The majority of telcos has to cope with an extremely complex, homegrown application landscape. Replacing this with standard packages greatly improves business flexibility, while reducing total cost of ownership. Prioritize business requirements over those of IT. The key to linking architecture to corporate-wide business

benefits is to always ensure that business requirements are the ultimate driver. The architecture transformation needs to be kept manageable and flexible toward changes in business priorities. This can be done via managed evolution, developing a high-level target end state together with a transformation plan consisting of several smaller projects (e.g., by domain or platform). To yield genuine value, an architecture transformation also requires a substantial shift in mindset. It is crucial to nurture the partnership between business and IT, rather than allowing IT departments to function in their default mode as delivery organizations and service providers. The designers or enterprise architects play an essential role in translating business language into

RECALL No 13 Leveraging Technology Fast forward to success: Managing next-generation IT architecture

41

IT requirements. Successful projects create followership, in turn broadening compliance with enterprisewide standards. Pursue seamless integration across the enterprise. Integration across the company is a vital driver of time to market and hence competitive advantage, as was highlighted earlier. A well-designed and implemented integration layer ensures maximum decoupling of functional blocks. Depending on the design decision, an integration bus can either ensure unified communication standards of atomic services or provide complex (compound) services with cross-domain business logic. Either way, it will provide transparency regarding the services provided by different domains and it will enable high reusability of existing functionalities. Integration of the data architecture across multiple domains is also key. All business activities create, modify, or influence information, making data the brain of the enterprise. Therefore, enterprise-wide information architecture management needs to be established as a self-standing discipline, yet run in close coordination with the application architecture. This is true for both master data, which is usually used by multiple domains, and transactional data, which flows with business processes and thus gains relevance in different domains depending on the status of the process. For example, telcos that attest to a much faster rate of growth after implementing high-performance customer lifecycle management only achieve this via state-of-the-art integration across multiple networks and database layers. Organize to maximize impact and compliance. As the case study (see text box) shows, teams with the appropriate authority and incentives are essential to keep transformation on track and ensure sustainability. A strong enterprise architecture management (EAM) function underpins this, leading the effort of designing and managing the transformation. One aspect of EAM is to define standardized artifacts: documentation on the current IT architecture, a blueprint of the target IT architecture, a roadmap for the IT architecture transformation, and a set of governance tools. Others are to set up a stable process with standardized roles and

responsibilities that are highly specialized. The EAM function needs to be empowered by the business and have sufficient resources and skills to monitor all of the major transformation projects. It should also be closely linked to IT delivery and operations to prevent it from developing ivory tower concepts and architectures. The most advanced methodologies and toolsets will be of little use if they are not applied in a day-to-day setting. Incentives need to be installed to motivate both the business and the IT sides to work along the new guidelines. Business designers need to understand the constraints and consequences of applying business solutions to the entire capability landscape. This can be done by creating close interaction with enterprise architecture managers and by using business cases. Application architects need to be convinced of the value of enterprise-wide management of IT assets to perform holistic optimization, otherwise they will remain wedded to local solutions. Tangible value for the IT delivery organization can consist of reusing components and services developed in other domains. Beyond this, proactively managing interdependencies between projects that would otherwise be discovered when systems fail to interoperate can prevent unpleasant surprises. Transforming a large telcos enterprise architecture management function to deliver maximum value is a Herculean task and multi-year effort requiring full buyin from the business side. There is no uniform panacea for success. But the impact on costs and business performance can be huge once the enterprise architecture moves toward a uniform blueprint with consistent management across domains. Tariff changes take days rather than months. Customers can be tracked across their lifecycle and targeted with optimally customized offers, while network utilization soars. As the McKinsey IT benchmarking survey has demonstrated, excellence in IT architecture is fundamental to efficiency and effectiveness, touching every aspect of a telcos business performance. Avoiding the pitfalls and heeding the pointers outlined above can greatly enhance the success of architecture transformation.

42

Duarte Begonha is a Principal in McKinseys Lisbon office. duarte_begonha@mckinsey.com

Klemens Hjartar is a Director in McKinseys Copenhagenoffice. klemens_hjartar@mckinsey.com

Matthias Roggendorf is an Associate Principal in McKinseys Berlin office. matthias_roggendorf@mckinsey.com

Olga Sterkhova is a Principal in McKinseys Moscow office. olga_sterkhova@mckinsey.com

RECALL No 13 Leveraging Technology Turning customer insights into income: The architecture is key

45

07 Turning customer insights into income:

The architecture is key

Customer lifecycle management can be a boon to both a telco and its customers, timing incentives perfectly to up usage and boosting the bottom line in turn. What is often forgotten is that without the right IT architecture, the best laid CLM plans will fail. In most markets even emerging the pressure on telcos to generate incremental value from their existing subscriber base is unprecedented. Revenues from new customer acquisition are slowing down, while competition intensifies by the day. The financial crisis in recent years has exacerbated the situation even further, with end users seeking out savings and potentially churning wherever they can. Customer lifecycle management (CLM) is one-to-one communication with the customer, generating a bond that can offer the user genuine value. CLM turns data into insights using a lifecycle perspective: telcos discover what the customer wants and deliver it exactly when needed, which enables them to win the price wars and transform commerce/business with better execution support. Leading-edge CLM techniques now engage thousands of customer microsegments with highly creative campaigns using below-the-line (BTL) multichannel communication, leveraging scale and speed. Doing this well can both boost revenues and provide a genuine competitive advantage, especially as CLM anchors institutional capabilities that are tough for competitors to observe, let alone replicate. However, many telcos are still failing to tap CLMs full potential. Their Achilles heel is their underlying IT

architecture. False starts and half-baked initiatives lacking adequate planning have left telcos with a spaghetti of fragmented, complex systems. To drive successful CLM, well-designed IT architecture needs to have scalable data marts, sophisticated analytics tools, a highly proficient microsegmentation engine, and a fully automated campaign management system. Plus, all elements must be capable of robust integration with each other and with network and content providers. This article outlines the requisites and suggests solutions. It begins by describing the CLM capabilities a telco needs to remain at the forefront of its rivals and then focuses on the optimal enabling IT architecture.

The building blocks of world-class CLM

In todays hypercompetitive market context, managing your existing customer base to extract higher revenues requires daunting scientific precision and reach. The first wave of CLM differentiates customers by value over their lifecycle, ranging from acquisition and up-sell to churn and win-back. The second wave incorporates multichannel execution, optimizing the match between channel and lifecycle stage. Incorporating retailers is a further vital element in the mix, adding yet another integration challenge. To stay ahead of the game, a telco needs to apply a unique set of capabilities across different fronts (Exhibit 1). Microsegmentation. The customer base is divided into hundreds to thousands of segments using static (e.g., historic usage) and dynamic triggers (such as change of

46

01

Telcos need to apply five core elements of customer lifecycle management Telcos need to apply five core elements of customer lifecycle management
100 - 10,000 segments

Usage aggregated and refreshed Static and dynamic parameters VAS and profiles (Internet,
ringtones, demographic, scoring, social network, etc.)
n ro io ic at M ent m eg

Offer and campaign creation

(including control groups) of lists and campaigns
m Cam an p ag aig em n en t
Com mun icat ion

Automatic generation Library of offers/grids

Dont guess, test!

Customer lifecycle management

Scientific marketing

Segmented dashboards ROI engine Analytics and models

Multiple channels Trigger-based

and networks communication

s lytic Ana

(USSD, IVR, SMS, etc.) Creativity

Integration with billing

Infrastructure

Data mart Team and processes

Scale and speed
SOURCE: McKinsey

balance, recharging, credits and adjustments, customer propensity toward a new model, and inactivity), instead of a handful of broad segments. Automated, test-driven campaign management. Campaign ideas are tested against a control group to find the optimal offers with a positive return on investment (ROI), creating a library of insights based on continuous testing. Using methods that are empirical, quantitative, and replicable is the key, rather than relying on intuition. Tangible proof that an idea works is also an invaluable tool for motivating the marketing team and soon the entire workforce, enabling the rapid transformation of mindsets and behavior. Communication with customers across multiple channels. Such channels include unstructured supplementary service data (USSD), SMS, or an interactive voice response (IVR) server. This contrasts with basing offers on a few standard channels (often SMS only) that can easily be duplicated by competitors. Advanced analytics. Optimal lifecycle management requires tools such as segmented performance dashboards and campaign ROI calculators along with data analytics and mining models that can be leveraged beyond CLM for overall marketing decision making.

Infrastructure for scale and speed. Running dozens of BTL tests/campaigns every week the best players may launch 20 to 30 a day requires state-of-the-art algorithms and automation. Results need analyzing in real time and quickly scaling out to the entire subscriber base. This is no comparison to the previous status quo, where only a handful of campaigns were launched per quarter based on tests with a select subscriber subset. Each of these CLM capabilities needs to have the underpinnings of a seamless, interconnected architecture to achieve peak performance and scale.

IT architecture for successful CLM

Optimal CLM design covers three architecture layers. The CLM business model layer shapes the underlying application landscape and infrastructure layers. The application layer comprises the following key CLM architecture components (Exhibit 2). CLM data mart designed for scale and speed. The CLM approach is hugely dependent on data availability. A ro bust data mart is the basis for all CLM applications that run on top of it, as it stores data on segments, offers, and campaigns. Data recency, scale, and integration flexibility are key features of successful CLM data mart design.

RECALL No 13 Leveraging Technology Turning customer insights into income: The architecture is key

47

02

Optimal CLM IT architecture comprises five components Optimal CLM IT architecture comprises five components
CLM component ARCHITECTURE LAYERS Business model Business operations Business capabilities Channels SMS IVR Portal Outbound dialer Call center Retailer

Microsegmentation engine Application landscape Applications and data IT integration platform Application layer

Campaign management

Analytics

Pre-existing internal/third-party transactional systems Billing Mediation Call data records (Re-)charging Third-party social VAS media Other internal systems

Integration layer Infrastructure Infrastructure services Information and communication technology

SOURCE: McKinsey

Communication

Data aggregation

CLM data mart PREPAID TELCO EXAMPLE

It is crucial that systems can capture sufficiently recent data. Elements that need daily updating include recharge behavior, balance, roaming status, inactivity, and minutes used to date. For other fields (e.g., campaign results), two- to four-day latency is good, but two weeks would be stale. Data like content genre or SMSto-voice usage ratio only require monthly updating. In terms of scalability, the data mart needs a cost-effective, extensible data model. The design should factor in future expansion in data volume (due to subscriber growth), data fields (incremental fields like value-added services), and data history. Large telcos with around 50 million subscribers may need 30 to 50 terabytes (TB) of data storage plus additional requirements for data staging and backup. Telcos are exploring technological advances using massive parallel processing platform architectures with clustered nodes to scale out costeffectively, columnar and correlation database management systems to reduce data volume by 50 to 70 percent, and 64-bit capacity to enhance computing speed. Implementing sophisticated, robust processes (e.g., periodic, proactive data volume and metadata/ontology management) provides additional benefits. The data extraction, transformation, and load layer needs to be flexible in terms of data integration, includ-

ing unified mediation systems across multiple network and database layers. The layer should be able to load billions of records within three to five hours from different distributed source systems. It should also have efficient compression techniques (e.g., a lossless compression ratio of around 40 to 90 percent). Another important feature is that it supports back, trickle, and streaming data loads. Telcos that have a well-designed data integration layer require only a few hours to load a data mart. The majority of telcos lack this capability, resulting in delays lasting several days. Microsegmentation engine to individualize multitudes into 100 to 10,000 segments. Such a tool should enable marketing analysts to create a segmentation grid using various dimensions (e.g., ARPU drop, usage behavior for voice, SMS, VAS, roaming, and social networks). In addition to this, it should perform cohort analysis in order to identify advanced segmentation patterns. The following aspects will prove important in designing an effective CLM segmentation tool. Data elements should be based on leading indicators. The focus should lie on collecting a few (i.e., 40 to 50) leading core and actionable indicator-based data fields (e.g., customer behavior, VAS, retailers), not the usual basic fields like revenues and usage.

48

Behavior analysis should include recent and dynamic customer behavior (e.g., an ARPU drop of over 50 percent in the last week, balance drop below the threshold, or disconnected caller tunes). Readily available dynamic behavior provides significant impact by allowing customers to be targeted at the right time. Customer profiling should include not just customer demographics, preferences (e.g., e-mail communication, handset capability), billing details, lifetime values, monthly margins, rate plans, and household relationships for postpaid telcos, but also scoring of social network metrics (e.g., centrality) along with psychographics (e.g., the customers next-purchase propensity). Advanced behavioral data includes data from content providers, such as interests (is the user a soccer or jazz fan?), use of ringtones, and Web searches to identify VAS and mobile advertising that could be appropriate. Variables that are frequently used for analytics and segmentation need to be preprocessed and stored in the data mart so they are easily available. Campaign management system to test multiple offer variants. Using customer insights from a microsegmentation engine to quickly design and execute campaigns and capture results is what differentiates high-caliber CLM. Efficient campaign management systems have near realtime customer behavior data and can auto-launch campaigns when users show certain behavior. The following features characterize good systems.

Tools should be capable of rapidly creating campaigns across all lifecycle stages. Good tools can select the target group with threshold size, generate communication scripts, schedule the campaign, and auto-filter opt-in/ opt-out. The system should also have functionalities to update and cross-reference with the exclusion lists/ hold-outs and track campaign history in a library. Automated campaign launch functions should be able to export target lists, make offers to target subscribers at predefined triggers, and alert/offer notification to call centers and retail outlets as well as subscribers. As mentioned earlier, good operators can launch 20 to 30 campaigns daily without any human intervention. Ideally, a campaign library stores the results of previous campaigns. This way, records are easily accessible if any successful campaigns from the past are relaunched. Analytics to market scientifically. A spectrum of tools from basic reporting to advanced analytics and data mining enable the design of a huge variety of tools. Three features that distinguish analytics tools can give a telcos CLM a competitive edge. Performance dashboards measure and visualize performance trends and opportunities by segment and by location. They focus on gaps (such as changes in revenue patterns) and decorrelation from mean reversion. Using such an approach, one operator was able to determine the root cause behind a revenue drop in one region. Around 80 percent of this decline was driven by

Revamping CLM data mart design to deliver value at scale and speed
One prepaid telco had a huge data mart that stored raw transaction data for one year (including three months call records online and the rest on tape). However, its systems had no automated functionality for deriving daily, weekly, or monthly data aggregates. The data mart was merely a collection of tables, without any standard reference data model, views, or business definitions. Data requirements related to a specific customer or a specific segment of customers had to be pulled out of the system and collated from multiple tables. This required extensive extraction time and manual work that limited CLM campaigns to three to four per day. This was even common practice among rival telcos as well. Ultimately, the telco decided to implement a CLM data mart over a year, comprising about 200 data fields including out- and inbound, select VAS, recharge, and denomination. The data mart was designed to cost-effectively scale out from 20 to over 50 terabytes over three years. This fully automated cutting-edge system allowed the telco to launch 2,000 to 2,500 unique offers, targeting over 5,000 microsegments, which delivered on the aspired 3 to 5 percent in incremental revenue lift.

RECALL No 13 Leveraging Technology Turning customer insights into income: The architecture is key

49

long-distance customers, the most of whom were buying their recharge coupons in just ten cities from three specific distributors. This insight enabled the operator to immediately focus on targeted corrective actions. Campaign ROI quantifies incremental revenue impact from an offer as measured against a control group. It also goes further, developing what-if scenarios to calculate hypothetical financial performance. Analytics and data mining are achieved through online analytical processing, drill-down, and slice-dice analysis along with data mining algorithms, including clustering, associations, and prediction. Particularly in the prepaid segment, churn prediction is poor as predictors are either too weak or come too late. However, using analytics to gain in-depth understanding of multi-SIM mobile or VAS user profiles can be very helpful. Communication to make creative offers via multiple channels. This integration layer enables consistent communication via multiple channels. Many telecoms operators leverage limited channels (typically SMS only) to communicate offers to their customers. Channel tests conducted at one operator showed that conversion rates for an offer increased by up to five times when voice interaction was used. An effective communication layer should integrate all inbound and outbound communication channels and relevant fulfillment applications, including the following. Interactive menu IVR should enable customers to accept offers based on simple-to-use interactive prompts. Providing a single IVR number for all customers and promoting this number to create awareness of segmented offers creates pull: customers proactively call the IVR to check for the best offer available for them. Telcos can also establish a link with resellers, providing them with information on the reseller offers and giving them incentives to push specific offers. This helps improve conversion rates, since resellers significantly influence the customers choice of recharge coupons along with other drivers. Using triggers can also prove very effective. One example: sending customers an offer using USSD either as a post-call notification or when they check their balance.

Integration with billing and fulfillment helps ensure customers receive special campaign benefits in real time. Any segmented campaign needs to be provisioned in IN (for prepaid) and billing systems (for postpaid) for the target customer base. In this way, only targeted customers reap the benefits from segmented campaigns. Also, such an interface ensures that any provisioning errors are corrected prior to sending out campaign communications to customers. Communication and integration should be based on service-oriented architecture/Web services to enable easy access via multiple communication channels, improve design flexibility, and reduce development and maintenance effort.

Why it is important to make the shift now

Operators who have developed the kind of advanced CLM capability described above have typically achieved an incremental revenue uplift of 3 to 5 percent and reduced churn by 5 to 10 percent. A robust integrated IT infrastructure significantly enhances the success of a telcos CLM capabilities. It drives scale: far more campaigns can be tested and launched, and customers contacted. It accelerates speed, ensuring the launch of new offers in less than a day via auto triggers. Consistency also soars as execution failures decline. Piloting, implementing, and rolling out the CLM IT infrastructure as outlined here requires significant effort, with initial deployment taking 12 to 18 months. For an operator with over 50 million subscribers, the capex needed for a fully integrated infrastructure can range from USD 3 to 10 million. But breakeven can be achieved rapidly within 12 to 18 months of deployment. It is also possible to achieve ROI of five to ten times the initial investment over three to five years. Being a first mover helps in fully capturing these benefits. A distinctive CLM infrastructure requires significant capability building in organization and in processes. Although a challenge, this is precisely why this, once achieved, is very hard for competitors to replicate. It can become a source of sustainable competitive advantage, substantial payback, and superior performance.

50

Neeraj Arora is a Practice Expert in McKinseys Silicon Valley office. neeraj_arora@mckinsey.com

Andr Levisse is a Director in McKinseys Singapore office. andre_levisse@mckinsey.com

Noppamas Masakee is an Associate Principal in McKinseys Bangkok office. noppamas_masakee@mckinsey.com

Samba Natarajan is a Principal in McKinseys Singapore office. samba_natarajan@mckinsey.com

Saket Pradhan is an Engagement Manager in McKinseys Delhi office. saket_pradhan@mckinsey.com

Kartik Sheth is an Engagement Manager in McKinseys Mumbai office. kartik_sheth@mckinsey.com

RECALL No 13 Leveraging Technology Transforming IT: An interview with Fernanda Torquati, Global CIO, Telefnica

53

08 Transforming IT: An interview with

Fernanda Torquati, Global CIO, Telefnica

Telefnica is one of the worlds largest telecommunications companies and has a presence in 25 countries and a customer base of over 273 million around the world. Its strongest footholds are Europe particularly in Spain, where it was founded in 1924 and Latin America. As of December 2009, it had a market capitalization of around EUR 56.7 billion. This company ranks third in the Eurostoxx 50. Mara Fernanda Torquati joined Telefnica Argentina in 2000, taking different roles in business support functions (including sales back offices, billing, collections) and in IT. In 2006, she became CIO of Telefnica Spain (wireline), and in 2007, as a consequence of the fixedmobile convergence, she combined this responsibility with the role of CIO of Telefnica Mviles (mobile in Spain) up to the end of 2008. She is now Global CIO of the Telefnica Group and leads their new subsidiary, Telefnica Global Technology. This was set up to unify the management and operation of its information systems worldwide. McKinsey had the opportunity to speak with Ms. Torquati about her broad perspective on IT/IS issues and the groups unique global IT transformation. McKINSEY: When you became CIO of Telefnica Spain in 2006, why did you decide to undertake an IT transformation program in Spain? FERNANDA TORQUATI: When I took on the CIO role, Telefnica Spains IT had reached high efficiency ratios, but it still had latitude for improvement in areas such as customer satisfaction, time to market, and availability. The company also began the fixed-mobile convergence process at more or less the same time. This required a

swift response from IT to enable converged functionalities. Addressing these challenges meant establishing a better dialog with the various business units (such as aligning priorities or using the same language), while also developing a more focused and structured IT organization the same went for our processes. To succeed, we also knew Telefnica needed strong partnerships with key IT vendors to ensure the requisite skills and capture additional efficiencies to finance the transformation. These aspirations and the changes they entailed demanded a structured transformation program with clear milestones and a bold vision of the target status. McKINSEY: What were the key steps of the IT transformation in Spain? FERNANDA TORQUATI: First, I would say it is vital to have an IT organization model with the right structure and skills. The structure needs to consolidate different IT islands across the organization and merge IT contracts. Decisive skills include key enabling functions such as client, business partner, and architectural management. I see these as the prerequisites of success. Models may differ depending on an operators needs, but a robust organization and the right skills are essential for navigating a successful transformation journey. The second step is to ensure production stability, reducing incidents to a minimum. Its wishful thinking to try to manage a large-scale transformation while fire-fighting at the production level. Third, operators need to find additional efficiencies to finance the transformation such as consolidating and

54

renegotiating IT contracts and ensure a proper balance between short-term priorities and the medium term. This step is key, otherwise day-to-day demands will drain away all the IT energy and resources, making it impossible to meet the transformation targets. Fourth, an operator should have a clear application roadmap fully aligned to business needs and priorities one that strikes the optimal balance between implementing new applications and upgrading legacies. Finally, you need to develop a continuous improvement mindset to drive optimization in both internal operations and business delivery. McKINSEY: You just touched on the balance between new applications and legacies. How do you think operators should transform their application landscape? FERNANDA TORQUATI: Thats a tricky question. As I have already mentioned, the key is to align the roadmap to business needs and priorities. The competitive context, market maturity, and quality of legacies can drive different types of roadmaps. I typically like to follow some guidelines when designing them. The first is to put rules in place to minimize implementation risk, like pursuing front-end (CRM/channels) and back-end (provisioning and billing) developments in parallel, or even promoting a big bang approach. I also try to follow the client process flow as much as possible CRM, provisioning, billing to uphold logic and consistency throughout all the processes. Its important to ensure business sponsorship, too. Without it, the best technical project is bound to fail. Additionally, you need to set up a project management culture, involving all your IT staff from day one. Service desks have to be on board before a new application goes live; development teams should plan how to manage the coexistence of two applications. In many ways, transforming applications equates to transforming the entire IT division. Managing the implementation time of the applications transformation is also crucial, ensuring short-term milestones aligned to a rock-solid endgame and constantly managing business expectations when quick wins fail to materialize, as they inevitably will. McKINSEY: What was the greatest challenge along the way? How did you overcome it? FERNANDA TORQUATI: Managing a transformation of this kind is a tremendous challenge. Obviously, ensuring a sufficient budget is a key requirement, but

that isnt enough. To achieve the transformation, CIOs have to make sure the key business leaders are involved, prioritizing and maintaining tight discipline during implementation. They also need to strike the right balance between day-to-day projects and new ones. This requires a dedicated and committed project management office. Highly skilled staff in just the right mix are vital as well architects, project managers, and IT planners. And you have to generate excitement surrounding the transformation process in the IT department, defusing any internal resistance or inertia. In order to manage all these elements simultaneously, it is critical that the transformation is seen as a company project, not simply the responsibility of IT. One other essential factor is that top management views the transformation as a key strategic priority. McKINSEY: In 2008, you took on the new challenge of Global CIO of Telefnica. What motivated you to accept this new position? FERNANDA TORQUATI: It was certainly a difficult decision. We were still in the midst of the transformation in Spain I felt my mission there wasnt finished. But being offered the opportunity to transform Telefnicas IT globally was very attractive. In the end, the group top managements aspiration to make IT Telefnicas first truly global function was what convinced me along with giving us the empowerment to make that happen. In this new role, I would be able to support the conclusion of the IT transformation in Spain, while triggering the same process in all other countries where Telefnica has a presence. McKINSEY: What do you believe are the key milestones in a global IT transformation? FERNANDA TORQUATI: A global IT transformation is a long journey. It depends to a great extent on the maturity and strength of the companys governance model and globalization strategy. To make it happen, you first need to share and agree on a common vision of what global IT and its benefits really mean. Second, an operator must ensure that it is backed by a proper governance model: CIO reports, an operating model, and a budget. Third, you have to define and align a transformation roadmap for all IT components. Then you demonstrate the benefits quickly to gain momentum data centers, infrastructure, and workstations are usually the frontrunners, since these IT functions require the least business process alignment. All these milestones are

RECALL No 13 Leveraging Technology Transforming IT: An interview with Fernanda Torquati, Global CIO, Telefnica

55

critical and the faster an operator can achieve them, the easier the journey becomes. But it will never be an easy ride. McKINSEY: What are the key success factors in global IT application architecture transformations? FERNANDA TORQUATI: I am still not sure there is such a thing as a global IT architecture transformation. Very few examples exist of global application architectures. To be honest, I dont think anyone has any relevant experience in the telecoms space. So many elements make it almost impossible to deploy a 100 percent global IT architecture: the size of Telefnicas global operations, the differing business needs and competitive contexts of our different operators, and the varying status quo of their legacy systems. However, there are some definite musts. Global IT should have a clear global application reference map, with preselected solutions that it can offer each country: best-of-breed customized developments, selected packages, and so on. Each operator should only deploy them if it really makes sense, though, and if the timing is appropriate. We mustnt forget that implementing a new CRM, provisioning, or billing system is only advisable if it would actually have a positive impact on the bottom line. Even if you agree wholeheartedly with the principle of implementing global applications, its still important to industrialize homemade applications and ensure some level of business process homogenization. All these aspects make a global IT application architecture transformation very challenging. McKINSEY: Your group has several operational bases. What level of IT integration do you believe is right? FERNANDA TORQUATI: At the end of our journey, I believe we will have a uniform global IT, while ensuring local demand management and an overarching country interface. But this wont happen overnight.

First , you have to find the right approach to globalizing IT. Decisions should factor in a number of practical criteria: after all, a global IT or application architecture should be a means of meeting larger business objectives, and not an end unto itself. McKINSEY: What are the greatest challenges of crossborder integration? FERNANDA TORQUATI: Several issues whether budget allocation, taxes, or legal and regulatory constraints represent major challenges that sometimes make us question if this is really possible. In the end, whats make or break is the groupwide determination to succeed. Achieving full alignment on making global IT transformation a reality from your local and global organizations is the key. McKINSEY: If you were starting now, is there anything you would do differently? FERNANDA TORQUATI: Looking back, I still believe we took the right steps both in Spain and globally. Obviously, there are always a number of small decisions that I would have made differently in retrospect. I would have been much faster and more straightforward about consolidating and integrating the different transformation processes accelerating the IT fixed/mobile convergence and unifying our BSS/OSS systems, while setting clear rules and boundaries. In addition, I would have been stricter about ensuring a deeper application transformation, forcing businesses to change their paradigm, and prioritize innovations around their online presence (channels, customer self-care services, etc.), even if they werent yet convinced. The key is to prevent short-term pressure from compromising your long-term strategy. Thats always hard, but I feel its particularly important. Fernanda Torquati was interviewed by Duarte Begonha, a Principal in McKinseys Lisbon office.

RECALL No 13 Leveraging Technology Appendix

57

Modeling lean in telecoms

A model factory for lean processes offers hands-on opportunities to experience lean in its original context, thus injecting cutting-edge best practices from manufacturing into the realm of telecoms. In 2007, Technische Universitt Darmstadts Institute for Production Management, Technology, and Machine Tools and McKinsey set up the CiP (Center for industrial Productivity). Its holistic transformation approach is based on three elements: Operating system. Develop and implement optimized end-to-end processes that create value for the customer with the least waste, minimal variability, and high flexibility. Management infrastructure. Align the organization to support integrated processes, implement performance management (KPI hierarchy and performance dialog), and set up an end-to-end infrastructure. Mindsets and behaviors. Understand leadership role modeling and management capability building. The CiP has developed training and continuing education programs covering the methods of lean management that target employees at all levels from technician to board member. It offers expert support solutions to individual company problems and serves as a forum to exchange theoretical and practical insights. Of the CiPs host of workshop modules, Current State Value Stream Mapping, Performance Management Metrics and Targets, and a full 20 others have significant telecoms relevance. Following each module, a manufacturing-to-telecoms knowledge transfer helps apply lessons from batch production directly to the telecoms industry. For more information, please contact Rainer Ulrich: rainer_ulrich@mckinsey.com

RECALL No 13 Leveraging Technology Appendix

59

News, views, insights

McKinseys Telecommunications Extranet is the gateway to some of the best information and most influential people in the telecommunications industry. The Extranet offers selected McKinsey-generated information that is not available in the general Internet. Extranet users have access to selected McKinsey articles on subjects relating to Industry & Regulation, Growth & Innovation, Sales & Marketing, Services & Operations, Next-Generation Networks, Technology & IT, Corporate Finance, Organization & HR, Corporate & Enterprise, and Equipment & Devices. Direct communication channels ensure that your questions and requests will be addressed swiftly. The site is updated weekly with new articles on current issues in the industry. McKinseys Telecoms Extranet lets you: Obtain exclusive information free of charge and use an Internet portal specifically designed for the industry Access cutting-edge know-how, interact with experts to gain new insights, and contact industry leaders Stay well-informed with daily industry news from factiva that you can tailor to your needs and interests. General information about the site is available at: http://telecoms.mckinsey.com Contact: telecoms@mckinsey.com

RECALL No 13 Leveraging Technology Appendix

61

Serving clients around the world

McKinseys Telecommunications Practice serves clients around the world in virtually all areas of the telecommunications industry and is part of McKinseys TMT Practice (Telecommunications, Media, and Technology). Our staff consists of individuals who combine professional experience in telecommunications and related disciplines with broad training in business management. Industry areas served include network operators and service providers, equipment and device manufacturers, infrastructure and content providers, integrated wireline/wireless players, and other telecommunications-related businesses. As in McKinseys work in every industry, our Practices goal is to help our industry clients make distinctive, substantial, and lasting improvements in their performance. The Practice has gained deep functional expertise in nearly every aspect of the value chain, e.g., in capability building and transformation, product development, operations, network technology and IT (both in strong collaboration with our Business Technology Office BTO), purchasing and supply chain, as well as in customer lifetime management, pricing, branding, distribution, and sales. Furthermore, we have developed perspectives on how new business models and disruptive technologies may influence these industries.

Telecommunications, Media, and Technology September 2010 Designed by Visual Media Services Germany Copyright McKinsey & Company, Inc.
www.mckinsey.com