Using data integration to drive down costs and increase profits in high transaction industries

sponsored by

Sponsors message
In 2006 Talend revolutionized the world of data integration when it released the first version of Talend Open Studio. After four years of intense research and development, and with solid financial backing from leading investment firms, Talend is now the recognized market leader in open source data integration. But Talend is not just the leading open source data integration vendor were also thought leaders, working at the cutting edge of next-generation data integration solutions. Our portfolio encompasses data integration (operational data integration and ETL for Business Intelligence), data quality and master data management (MDM) technology. And unlike the quickly consolidating traditional vendors offering proprietary, closed solutions, Talend offers a completely new vision of data integration. Not only do we shatter the traditional proprietary model by supplying open, innovative and powerful software solutions with the flexibility to meet the needs of all organizations, but were also making data integration solutions affordable for organizations of all sizes and for all integration needs. Telesperience research shows that modern businesses are facing rising demand for data integration. Coping with this demand means that manual integration is no longer a viable option. Yet specialist data integration tools are often too expensive and simply dont offer the range of functions or the flexibility required by todays data integration specialists. If youre one of those companies wondering how youll deliver the data integration projects your business is demanding on time and to budget then maybe you should be looking at Talends solutions. To find out more about what we can do for you visit www.talend.com, where youll find a wide range of information about our solution, a library of thought leadership and data integration resources and, of course, open source data integration, data quality and MDM software which you can download today.

Yves de Montcheuil VP Marketing, Talend January 2010

Telesperience Research 2010

Telesperience Issues Paper: Using data integration to drive down costs and increase profits in high transaction industries

Summary

Data integration is often seen as merely a technical discipline; but this ignores the vital role it plays in helping enterprises achieve their business goals. It is essential that business managers, as well as technical staff, understand how better data integration can help deliver against key commercial goals, such as helping the organisation become more efficient, agile, innovative and customer centric. Business managers should think of data integration as information integration, which supports better business decisions, makes the business more operationally efficient, and helps deliver an improved customer experience, service innovation & differentiation. Well-integrated data also helps organisations identify new markets and new business opportunities, and improves their ability to cross-sell and upsell to existing customers. This paper outlines why data integration is an important weapon in an enterprises competitive arsenal. It explains how integrating data can help companies minimise their operational costs and maximise their opportunities. It also presents the findings of primary research Telesperience has conducted in the telecoms, banking, computing/IT and government sectors. Key findings include: the main drivers for DI initiatives are improving the customer experience (64%) and reducing costs (55%) a pattern of high and increasing demand for DI. Twenty-seven per cent of companies said their key goal in 2010-11 was simply to cope with increasing demand for DI from the business 46% of companies said their current DI tools do not fully support what they want to do 72% per cent of companies told us their DI projects often over-run on time, budget or both poor DI leads to higher direct costs. Fifty-five per cent of companies we spoke to said it led to higher manpower costs and 37% said they aimed to reduce the manpower required to support DI projects in the next 24 months DI is a strategy for lowering costs and improving business performance. Poor DI leads to huge and often poorly-understood indirect costs. Forty-six per cent of companies told us that poor DI is preventing them from taking advantage of new commercial opportunities DI costs are high and 73% of companies said their costs were either too high and could be reduced, or poorly understood and often more than anticipated. DI costs are driven up by common project management challenges such as poorly-defined business requirements. Manpower costs are also high and firms have identified increased automation as a desirable way of reducing overall costs. This suggests increased use of DI solutions in the next two years

Definitions and key concepts

Data integration (DI) is not just a technical challenge, but also a key enabler of many of todays strategic business goals, such as reducing costs, improving the customer experience, innovating, managing the business more effectively and complying with regulation or legislation. DI also helps businesses identify new opportunities and increase wallet share from existing customers. It is therefore vital business managers understand why it is necessary, exactly what it is & what it encompasses, and the business benefits it delivers. Why is data integration needed? DI is necessary because of the way information systems have been architected and grown over the last 20-30 years - a situation exacerbated by rapid organisational expansion, tactical investment strategies and M&A activity. Typically, this has resulted in enterprise data being fragmented across multiple IT solutions and their corresponding data silos. At the same time, enterprises now want to use more data, and a more diverse range of data, for both operational and strategic purposes. This is driving the need for DI particularly in larger organisations. What is data integration? Defining DI may seem obvious, and practitioners may have an intuitive understanding of what it involves, but in fact it is far harder to define than it might first appear. This is because the term implies different things depending on the context and upon those using it. In essence, DI involves combining different sets of data to provide a unified view of all the relevant data. It is a concept that is implemented through a combination of methodologies and technologies, and it encompasses database migration/upgrading, application migration/consolidation, operational DI and integration to support data analytics. Operational DI also facilitates initiatives such as customer data integration (CDI) aimed at improving the customer experience and the profitable operation of the company. Analytical DI is used to support business-level initiatives such as business intelligence. What does data integration encompass? There are a number of approaches to delivering DI and these can be used individually or in combination, as appropriate: data consolidation involves collecting data from multiple data sources and consolidating it in a persistent data store. This may involve, for example, migrating data from multiple existing datasets and consolidating these into a single persistent dataset as part of an application migration or consolidation initiative. A number of technologies can be used to assist with data consolidation, including extract-transform-load (ETL) tools and so-called 3G data migration tools data federation provides a unified view of an organisations data through a single interface, enabling disparate data sets to appear as a single homogeneous data set to the user. This is also known as enterprise information integration or EII data propagation involves replicating data from different sources in different locations and encompasses enterprise data replication (EDR), database log scrapers and change data capture (CDC) tools. Enterprise application integration (EAI) technology supports the integration of application systems, enabling them to exchange data using standard interfaces data access uses search capabilities to increase the accessibility of data. This is also known as enterprise information access or EIA. A key concept in DI is legacy data, which can be structured or unstructured, and includes paper documents, database records (sometimes stored on mainframes), spreadsheets and word processor documents. Other technologies that are frequently employed in DI initiatives include discovery and

Telesperience Research 2010

profiling tools, metadata and master data management software, reporting tools and data quality tools. Few tools can operate in multiple DI modes or have the functional scope to cover a wide range of tasks. Typically, organisations still employ a combination of tools to perform complex DI tasks, and hand-coding is still very common. Specialist tools are, however, being transformed into suites and DI solutions often as a result of M&A activity. Firms have recently begun moving from batch-oriented DI to realtime or online DI. Realtime DI is not ubiquitously appropriate and can impose a significant cost and performance overhead. Some applications and decision-making processes will only ever need to operate in batch and so realtime DI is unnecessary; others will benefit significantly and deliver higher commercial value through access to the most up-to-date data. Change Data Capture tools, for example, ensure that target datasets are updated as changes are made to data sources, so realtime queries are based on the most current and accurate data. EAI technology is often used for realtime operational business transaction processing.

How does poor data integration drive up costs?

direct costs for example, due to higher hardware and software licences, and the requirement for greater manual effort indirect costs for example, opportunity costs, costs arising from poorer or slower business decision-making, or as a consequence of negative impacts on customers (resulting in higher rates of customer churn and the costs incurred from dealing with complaints).

Poor DI increases the costs for businesses in a number of ways (see Figure 1). These costs are either:

Telesperience research conducted amongst large enterprises in the telecoms, banking, computing/IT and government sectors found that the most commonly cited costs of poor DI were operational inefficiency (73% of respondents), opportunity costs (55%), and higher labour costs due to the requirement for manual intervention (55%). The fact these costs are recognised as significant is not related to their size. Costs arising from higher cost operations and the use of extra manpower may be significant and quantifiable; however, although harder to quantify, the cost of lost opportunities may dwarf these figures. It is important that indirect costs are taken into account when considering the cost:benefit analysis of a DI exercise, and the ROI of a DI tool or project. Figure 1 Scenario 1: poor data integration increases costs & leaks value from the organisation

Costs up Slow and sub-optimal decision making Poor investment decisions Risk of non-compliance Poor competitive positioning

High hardware costs High software costs High operational costs High opportunity costs High customer care costs

Sub-optimal commercial experience

In the red: the inefficient enterprise

Poor customer experience

Slow & inaccurate service delivery Inability to adapt or innovate Slow time-to-market Too much manpower required

High churn rates Reactive customer care Poor/ineffective marketing Slow time to resolve SLA breaches Operational inefficiency

Key data integration drivers

Most common costs of poor data integration

Source: Telesperience 2010

Driving down costs using better data integration

Current data integration drivers

Sixty-four per cent of organisations told us that the main driver in their organisation for DI was an initiative to improve the customer experience. Cost reduction (55%) and compliance (46%) were the next most commonly-cited drivers (see Figure 3). Going forward, organisations said that in the next 24 months decreasing human effort (37%) and coping with increased demand for DI from the business (27%) were their main goals. These findings present a pattern of ongoing even overwhelming demand for DI. Thirty-seven per cent of organisations told us they recognise the need to remove manual intervention from the process and a further 18% told us that speeding up project timescales was important to them. The consequence is likely to be increased use of technology to speed up, automate and make DI projects cheaper, as well as to help organisations cope with the sheer demand for DI from the business. The telecoms, computing/IT and government sectors were particularly concerned about coping with increasing demand for DI; banking was the only industry that cited reducing risk as a key goal although we believe all organisations should be seeking to reduce risk. Figure 3 Drivers of data integration projects
Current DI drivers 2008-9
Other Decrease amount of human effort Decrease risk Compliance Improving the customer experience Modernisation

Future DI goals 2010-11

Decrease cost Decrease time taken for data integration projects Cope with increasing demand for DI from the business 10% 20% 30% 40% 50% 60% 70% 80% 5% 10% 15% 20% 25% 30% 35% 40%

Cost reduction

Source: Telesperience 2010

Using data integration to deliver commercial success

Integrating business data creates a wide range of benefits for an organisation (as shown in Figure 3), helping to deliver commercial success and competitive differentiation. The potential of DI will only be realised, however, if businesses expertly manage the DI process. The huge scale of potential savings and commercial benefits is such that business managers should be championing DI projects, helping set their goals and monitoring their delivery. It is vital that DI receives the support, buy-in and involvement of the business if it is to be successful. Fifty-five per cent of companies we spoke to said poorly-defined business requirements were raising their organisations DI costs. If the businesss needs and objectives are not clearly captured at the onset, and monitored throughout, then the project is highly unlikely to deliver against the organisations business goals. This is a common cause of projects failing to achieve their objectives or running over on cost or time. In this section we look at three main areas that businesses can focus on to deliver the commercial benefits we have outlined and to reduce costs: reducing the cost of the DI project itself reducing the direct costs of poor data integration reducing the indirect costs of poor data integration.

Telesperience Research 2010

Figure 3 Scenario 2: well-integrated data is a vital component of commercial success value is retained by the business
Hardware costs optimised Software costs minimised Operational costs minimised New opportunities identified and acted upon Lower customer support costs

Fast and effective decision making Better investment decisions Improved compliance Better competitive positioning

Cost optimisation

Optimal commercial experience

In the pink: commercially successful enterprise

Optimal customer experience

Fast and efficient service delivery Supports adaption and innovativion Fast time-to-market High level of automation

Operational efficiency

Churn rate lowered Proactive customer care Personalised marketing Fast time to resolve Improved performance against SLAs

Key data integration drivers

Addresses most common costs of poor data integration

Source: Telesperience 2010

Reducing the cost of the DI project

More than 55% of companies we spoke to said their DI costs were too high and could be reduced, and a further 18% said the costs were poorly understood or often more than anticipated. We asked companies what was contributing to higher DI costs (see Figure 4). Most of the factors cited are well understood; most are avoidable; all can be mitigated. Yet these same factors continue to cause problems repeatedly. The best performing companies have taken steps to manage these issues. They understand the need to invest in good project management, see DI as a valued skillset, and recognise that the business should be intimately involved in DI projects. Companies that undertake only a relatively small number of DI projects are understandably particularly vulnerable to over-runs, because they lack the experience to avoid well-understood pitfalls. However, even experienced companies continue to make the same mistakes and run into the same problems. This is exemplified by the fact that only 28% of companies told us they were confident they could deliver projects to time and budget, while 72% admitted that they often over-run either on time or budget (or both). Figure 4 Factors driving up data integration costs
Business requirements not well defined Over-estimation of data quality Repeating or multiplying effort, lack of re-use Delays/lengthy project cycles Internal politics or conflict Lack of support/conflict with vendors SIs consultants etc Lack of expertise Poor project mgt Tools do not fully support what is needed Cost of tools/hardware Manual effort 0% 10% 20% 30% 40% 50% 60%

Source: Telesperience 2010

Driving down costs using better data integration

Optimisation checklist A: delivering more efficient DI projects A1. Capture business requirements accurately at the start of the project A common cause of project delays and failures. The business needs to be represented in the project team because technologists cannot be expected to understand business priorities. Business priorities need to be monitored throughout the project particularly if it is expected to be lengthy to ensure that the project continues to deliver against these needs as they change and evolve. Achievable: Cost savings: $ $ Other benefits: faster project times, more business-oriented results, greater buy-in from business users A2. Dont overestimate the data quality A common cause of project delays and unplanned costs. Dont assume that data quality is good or that you necessarily know where the data you need is. Put time and money into the budget to deal with data quality. Its no good integrating data if that data is poor quality, so addressing data quality is an important element in the success of your data integration initiative. However, dont get too carried away with data quality it is not practical or cost-effective to have 100% accuracy. Achievable: Cost savings: $ $ $ Other benefits: better end results, more accurate project planning A3. Plan an achievable project dont overscope Data integration projects need clear scoping and teams need to be mindful of scope-creep. Sometimes delivering against business goals can mean changing the priorities of the project; but scope-creep will cause delays and delays cost money because business priorities are likely to shift the longer the project takes to complete. The best performing companies with regard to DI usually have the fastest and most targetted projects. Achievable: Cost savings: $ $ Other benefits: faster project times, faster time to benefits, buy-in from business users A4. Sort out the politics Our research demonstrates that internal politics and conflict are a major source of costly over-runs. Do not underestimate the potential for conflict and ensure that you have buy-in from all stakeholders. Put effort into communication, requirements capture and achieving buy-in. Monitor the situation throughout the lifetime of the project. Do not allow the project team to become isolated from the business. Achievable: Cost savings: $ $ Other benefits: better results, boosts uptake/buy-in, less risk of over-runs A5. Review your use of DI technology and pay attention to software licensing and other DI tool costs Many organisations will have redundant or duplicated DI tools and may still be paying out for software licences, maintenance and so on. For each type of DI select a tool and standardise on it if possible, removing redundant tools. Consider using a multi-mode solution that can be re-used for other projects, and consider whether open source technology is a viable alternative. Around two-thirds of companies already employ at least some open source DI technology1 and its use is increasing. By reviewing your use of DI tools it is possible to save considerable sums while also benefitting from state-of-the-art technology that will help you achieve better results. Achievable: Cost savings: $ $ $ Other benefits: access to state-of-the-art technology reduces requirement for manual coding/intervention
1 Usage

Landscape: Enterprise Open Source Data Integration, Talend

Reducing the direct costs of poor data integration

Poor DI raises operational and capital costs directly by inflating hardware costs, software costs and manpower costs. With the increasing emphasis on cost-optimisation, and the current pressure on capital and operating budgets, this is wasted money that could be redeployed for innovation or cashed in to fund lower pricing or higher dividends. Moreover, DI projects themselves can result in inflated costs to the enterprise due to licensing terms and the requirement for high levels of manual effort. Hardware costs Twenty-seven per cent of organisations told us that poor DI was increasing their hardware costs. Hardware costs can be inflated due to:

Telesperience Research 2010

storing duplicated data some data duplication may be necessary for operational or performance purposes, but reviewing legacy infrastructure often reveals datasets that are duplicated for no good reason. This increases storage costs and also leads to operational inefficiency aging hardware infrastructure legacy hardware is usually more costly to maintain than the modern equivalent, while also having lower performance. The trigger for some data migration projects is to reduce the cost of legacy hardware platforms. In our survey, 18% of companies said DI issues were preventing them from modernising their infrastructure, and this was driving up their costs use of high-end hardware companies such as Google have changed the rationale of hardware strategies. Google famously uses commodity hardware, and does not even invest in the best commodity hardware, but rather that which provides the best cost per query. This raises a fresh challenge for other firms to match this cost:performance profile which, in turn, means they will require a DI solution that can support this type of infrastructure refresh under-utilisation of hardware in the past there was a tendency to deploy hardware for each application, regardless of whether there was existing capacity elsewhere within the organisation. This leads to a higher than necessary cost profile, which can be significantly reduced by consolidating data and applications onto a smaller number of hardware platforms. Adding on new data centers, which can cost up to USD100 million each, is not a viable option when server utilisation is low (industry estimates suggest server utilisation is currently only around 15% on average, and that around 10% of servers are unused although they are still consuming power, cooling, space and maintenance resources). It may not be possible or desirable to build new data centers due to space restrictions or the need to comply with green targets such as reduced power consumption. Strategies such as migrating and consolidating data on a modern, lower-cost platform, employing virtualisation and cloud computing are all being used to lower the cost of hardware for organisations. DI technology is an enabler of all of these initiatives.
Case study: potential for saving money on hardware The scope of savings on hardware is entirely dependent on the legacy environment, but can be considerable. An example from Sun Microsystems exemplifies this potential. Sun calculates that upgrading from a legacy system to Sun-optimized Oracle CRM on Sun Fire T5440 servers delivers hardware savings of USD1-5 million within five years, ROI in the range of 700-1900%, and power and floor space reductions of up to 90%. These results are based on tested and verified benchmark results published at http://www.oracle.com/apps_benchmark/html/white-papers-siebel.html

Software costs Most organisations are paying too much for their software. A review often reveals: software that is no longer being used or is underused, but is still being supported and maintained. Typically, 75% of the lifetime cost of software derives from maintenance costs, which means managing it and understanding lifetime TCO is of vital importance licensing costs that have risen exponentially due to the pricing model being used. Many organisations still only consider the initial purchasing price rather than TCO and are shocked when they realise how much software is costing now their business has grown licensing costs that are inflated because the IT department bought more than it needed (eg to get volume discounts or because the organisation has subsequently contracted in size) inflated costs for licensed software where lower-cost viable alternatives exist. Many IT users buy into a brand. Where such software is delivering significant added value then there may be a business case for paying more; but sometimes the software is delivering no more added value than a lower cost or open source alternative departmental purchasing strategies that fail to leverage volume economies bloatware costs where IT departments buy far more functions than they need or are used. This also adds to hardware, training and maintenance costs.

Driving down costs using better data integration

After reviewing software spend, the logical next step is a rationalisation programme to reduce the cost of software to the organisation, and this is where DI has a key enabling role to play in helping companies migrate critical data to new applications or to support application consolidation. However, the cost of a data migration project can also be inflated by software licensing. This is because organisations may select a tool with a poor price:performance ratio, unfavourable licence terms and so on. Tools are often bought on a case-by-case basis, rather than organisations properly assessing and investing in a tool that is reusable, presents a good price:performance proposition and will have a favourable TCO in the long term. In fact, 27% of organisations we talked to said that the cost of DI tools significantly impacted on their project costs.

Case study: potential for saving money on software Specialist companies exist to help organisations understand their usage of software and hardware as a precursor to a data migration or consolidation initiative. For example, IT discovery vendor Tideway, recently acquired by BMC Software, says that one of its customers, a major European power utility, used the insight it provided to reduce Oracle licence renewal exposure from GBP3.2m to GBP850,000 - simply by getting a better understanding of the licence position. By capturing an accurate view of their hardware assets and software licences, firms are able to initiate end-of-life programmes and consolidation initiatives, delivering additional cost saving opportunities.

Manpower costs Many companies are faced with higher than necessary operational costs (OPEX) because the legacy software/platforms they use require considerable manual intervention and the skills needed to maintain them are scarce and expensive. This also raises the risk profile for an organisation. The requirement for excessive manual intervention can result in a further hidden cost: most companies have limited resources, so valuable skilled staff are often tied up just keeping legacy IT running rather than being used to drive innovation or add value. This slows the innovation and renewal cycle, and means organisations are not maximising the value of their skilled staff.
Case study: using opensource technology to save money on DI projects and software According to Talend, a vendor of opensource data integration tools, opensource technology can help organisations save money by automating tasks that previously required manual scripting. A recent survey by Talend found that many (55%) large enterprises are still invoking manual scripting to keep information flowing across the organisation. Talend says use of opensource technology is on the rise, with 31% of companies reporting they combine commercial applications, opensource solutions and database utilities to meet their data integration needs. Telesperience research suggests that 18% of respondents are currently using opensource technology within their business, 9% are actively investigating it and 55% are interested in using it or have not yet ruled out the possibility of using it. But how does opensource technology save organisations money? Predictable pricing eg Talend bases its pricing on numbers of users not CPUs or data volume Use of scalable but low-cost hardware Low R&D costs since many innovations/functions are developed by the community there is a re-use of competencies Low start-up costs, making it much cheaper for one-off projects Talend says that its tool has a much wider range of competencies than many rivals, because of the constant, active development and innovation by community members. See: Usage Landscape: Enterprise Open Source Data Integration, Talend

In terms of the migration process itself, 55% of companies told us their data migration projects involved excessive manual effort, and this was a major contributor to DI costs. Rising demand for DI from the business, combined with limited budgets, means that throwing more people at the migration is increasingly unviable. The reason for so much manual effort was revealed by the fact that 46% of companies we spoke to said that their data migration tools do not fully support what they need to do, and this results in more manual intervention than is desirable and ultimately higher costs. This often occurs because of the use of specialist DI tools rather than firms employing flexible DI solutions that can operate in more than one DI mode.

Telesperience Research 2010

Optimisation checklist B: cutting direct costs from poor data integration B1. Pay attention to licensing costs Data integration software costs can stack up quickly. Pay attention not just to upfront costs but lifetime TCO when selecting a data integration tool. Typical tools range from USD200,000 to USD500,000 for licensing plus another USD50,000 to USD100,000 for maintenance. Consolidating and standardising tool use and investigating more cost-effective options (such as opensource technology) can quickly save large sums. Application consolidation can lead to huge savings in licensing costs, and this is enabled by DI technology. You are advised to cut out bloatware due to over-specification of functions. Look for applications that enable you to add on modules or functions when you need them. Pay attention to licensing terms are you buying more licences than you actually need? Will costs rise out of control if your business grows? Achievable: Cost savings: $ $ $ Other benefits: easier/cheaper training and management B2. Pay attention to storage costs Gaining cost efficiencies from modern, higher performance hardware is only possible if you can switch off legacy hardware. DI over-runs result in paying out for overlapping licensing, leasing, maintenance and so on, which can quickly rise to millions of dollars of unnecessary extra cost. By investing in a reliable, low-risk DI solution you will be able to deliver migrations on time or even before time. This will save you far more than it costs, because the faster you can switch off legacy, the faster you can get to the benefits of the new infrastructure. Achievable: Cost savings: $ $ $ Other benefits: less manual intervention, less downtime B3. Automate as much as possible Often costs rise with data integration because companies choose the wrong tool for the job. This means they end up having to write manual scripts and manage far more of the process manually. They may even decide not to use a tool and integrate manually. This drives up costs, makes over-runs more likely and is far less efficient. It is really important to understand the type of integration you are trying to achieve and select the right solution for the job. Third-generation data integration tools are more flexible, more configurable and more functional than previous generations and suitable for more complex data integration scenarios. Achievable: Cost savings: $ $ Other benefits: faster time-to-benefits, minimise use of scarce and expensive skilled resources B4. Do you need all the data? Costs rise because organisations store duplicate data, data they no longer need and so on. There is a cost associated with keeping, integrating and migrating data records, and it is essential that you understand what that cost is and use this to help decide which data you need to keep. Taking the approach of moving it and then deciding what to prune results in higher migration costs, extended timescales and the temptation to never consolidate/prune the data. Prune back what you dont need before you migrate or integrate. Achievable: Cost savings: $ $ Other benefits: compliance, faster DI projects, higher data quality B5. How flexible is your DI tool? Ask yourself: could you save money by being able to re-use the tool for further projects? Can the tool operate in more than one DI mode allowing you to get rid of other specialist tools and save on licensing/maintenance? Is it flexible enough to accommodate changes in business objectives part-way through the project? Achievable: Cost savings: $ $ Other benefits: faster project times, greater chance of successful project

Reducing the indirect costs of poor data integration

The indirect costs of poor DI may be recognised in the best-performing organisations, but even then are often not fully understood or captured. Indirect costs may be so large that it makes DI a no brainer for the organisation as a whole, but because responsibility and consequences are fragmented between departments the importance of DI to the business may not be recognised. For example, inflated customer support costs and a poor customer experience may be bad for the company, but if this is caused by the poor performance of one department (such as poor fault management - itself due to not having access to the right data at the right time), then the affected department (e.g. customer support) may not be able to influence change. Fixing the problem might also not be a high priority for

Driving down costs using better data integration

the department causing it. This is why the business consequences of poor DI need to be considered at an enterprise level and action prioritised according to the commercial impact. Failure of business-level analytical initiatives, such as business intelligence (BI), is often not due to the software itself, but instead is rooted in poor DI or data quality. These projects promise great improvements in business performance, but there is little point in spending large sums on analysing data if the data is poor quality, incomplete or inaccessible. The most obvious indirect cost of poor DI (46% of organisations said this was a big issue for them) is that it prevents the business from taking advantage of new commercial opportunities and increasing their market share. This results from supporting systems not being in place to support new service rollout; or because vital customer data is not available for analysis to inform better decision-making, new product design, promotions or pricing strategy.

Case study: minimising opportunity costs and opening up valuable new revenue streams When BT decided to migrate its FeatureNet customers - large corporates using BT VPN services and VAS - from a custom-built legacy solution for billing, tariff management and dial plan management to Convergys's Geneva solution it knew the migration would be complex and potentially risky if not managed well. The legacy solution supported some of its largest and most valuable accounts, a significant number of complex, multi-layered tariffs, as well as multi-tiered and multi-layered discount schemes. Using modern DI technology de-risked the project and enabled it to be completed in around six months. This meant BT could launch innovative new services to its FeatureNet customers 12 months ahead of schedule, delivering significant new revenue streams which would have been lost or delayed if the migration had failed or had been late.

Poor DI also drives up churn - primarily because it leads to a poor and inconsistent customer experience. Twenty-seven per cent of companies told us that poor DI resulted in increased customer care costs, and 9% said it was leading to higher churn rates. In our view, the recognition of these effects is likely to be underestimated.

Optimisation checklist C: cutting indirect costs from poor data integration C1. Understand the total effect of sub-optimal DI It is essential to take a business-level view of how poor DI is raising costs for your business. You need to understand not just the direct costs to the data owner, but how this is impacting on customer care, churn rates, market positioning and your ability to rollout new services, prices or promotions. Also understand how poor DI will lead to the failure of expensive and desirable business-level initiatives such as BI projects. Be aware, however, that whatever figure you come up with is likely to be underestimated, since understanding some of these costs is difficult. But capturing the impacts and estimating the cost is important to building the business case for improved DI. Monitoring the ROI post DI project is achievable in the form of eg reduced churn rates, lower customer care costs and new revenues from services not previously supported. Achievable: Cost savings: $ $ $ Other benefits: better corporate governance, improved commercial performance C2. Dont underestimate risk and compliance issues Depending on your industry and locality, a range of legislation and regulation will apply to your data and corporate governance. One of the biggest costs for some organisations is brand damage due to poor publicity. Poor data integration may mean you cannot comply or cannot prove compliance. It may result in embarrassing press stories, which in turn can have a negative impact on stock prices. If you operate in an industry where brand reputation is important, then DI should be high on your list of action points due to its potential to cause embarrassment and loss if not handled well. Achievable: Cost savings: $ $ $ Other benefits: maintain brand values and stock price, avoid regulatory or legal action

10

Telesperience Research 2010

Conclusion
Data integration (DI) plays a key enabling role in many important operational and strategic initiatives. The costs of poor DI and the business benefits of delivering good DI are such that it should be high on business managers agendas. DI cannot be seen as merely a technical task, because without the intimate involvement of the business in DI projects they will not deliver to their potential. The demand for DI will continue to rise in large enterprises and this is why selecting the right solutions is essential it is no longer viable to manage with a combination of spot tools and manual integration. Throwing bodies at the problem is also not an option. Companies should consider the full range of options available to them and investigate 3G DI tools that offer a much wider range of functions, are more flexible and can operate in more than one DI mode. Companies that are able to deliver DI projects quickly, reliably and at low cost will out-perform their rivals, because they will be able to exploit new technologies and insights to gain competitive advantage and reduce their costs.

Acknowledgements
The authors would like to thank all those companies and individuals who helped with our DI research and generously gave their time and expertise. In particular, we would like to thank Talend, who provided sponsorship to fund this paper, Sun Microsystems, BT and BMC Software.

About Telesperience
Telesperience is a UK-based telecoms analyst firm focused on how technology impacts both the commercial and customer experience. It is wholly-owned by Babworth Ltd, a provider of research, publications and writing services to the global Internet, Communications and IT markets. The scope and focus of Telesperience is as follows: the commercial telesperience to analyse how key IT technologies impact on telecoms service providers businesses the customer telesperience - analysing how key IT technologies impact on the end customer experience.

Telesperience was founded in 2008 by an experienced team of telecoms IT analysts who wanted to provide a more convergent view of the telecoms market, focusing on business and customer issues. We consider where the problems lie with legacy technology, and how companies can transition to provide a more positive telesperience for their customers and a more profitable business for themselves. Telesperiences open source research programme relies on the goodwill of companies who fund research in order to make it free at the point of delivery. We endeavour to ensure that our research remains objective and independent the steps we take to do this are outlined on our website, but the most significant is using experienced and respected analysts who have a track record within our industry. Report sponsors are always acknowledged, so readers are aware who is funding the research programme. Find out more about Telesperience at www.telesperience.com and www.microsperience.com.

11