Published on June 20, 2013
Thought Leadership White PaperIBM Transaction Banking June 2013Global platformreplacement:practice, issues andrecommendations
2 Global platform replacement: practice, issues and recommendationsExecutive summaryGlobalised organisations, both financial and industrial, arefacing the problem of modernising ageing and fragmented ITplatforms at a time when globalisation has a direct influenceon customer and regulatory demands, making this an eventougher challenge to overcome.This paper, based on fresh IBM research, examines how suchorganisations are tackling the problem when the only viableway forward is replacement. It covers both the architecturalchoices they make and the long-term implications that result.We found the following from our research:•• The investment and time required to make a successfulprogramme is substantial and often underestimated.•• The costs outweigh the business case, bringing a challengeto any organisation that wants to undertake them.•• There is no dominant pattern of design thinking,organisational engagement or programme execution.•• Some organisations have found it too difficult to delivera replacement programme in full.However, from the research we are able to do the following:•• Propose a set of prerequisites that significantly increasethe chances of delivering a successful project.•• Identify a set of challenges for the industry to overcomein order to realise the benefits from the investment.IntroductionRetail and corporate banks were some of the first andmost enthusiastic adopters of computing systems in thelatter half of the twentieth century. The architecture andimplementation of systems running the core operations ofmost major banks can be traced back to then. Since theseinitial implementations, their maintenance and enhancementrepresent huge investments and the embodiment of vastamounts of organisational knowledge.These banks are now facing unprecedented pressure tochange, from organisational costs, constantly growingregulatory demands, rapid evolution of technology withinhost system environments, ‘24/7 anywhere’ access to servicesand customers demanding improved service and information.Many organisations are considering how to respond to thispressure, and for some, renovation and componentisation ofthe current systems are judged to be the best ways forward.For others, the radical step of core system replacementappears to be the only way forward.Given the size and complexity of the challenge, IBM hasconducted research to establish how best to set up such coresystem replacement projects successfully.This paper is particularly directed towards IT and businessexecutives whose organisations have chosen replacement.In the context of the financial services industry, this chieflymeans Tier 1 banks with a global presence.It examines how they, as well as global players in otherindustries, are responding to the challenge of replacingtheir existing global platforms. For the majority of suchorganisations, this represents the first attempt to replacelegacy platforms that have evolved over the last 30 years.It addresses the following questions:•• What are banks and other major global organisationsdoing, and why?•• Are the large budgets and long timescales really necessary?•• Which approaches are most successful and where areorganisations struggling?•• What can be done to increase the chances of success?Contents 2 Executive summary 2 Introduction 3 Research approach 4 Discussion 9 Understanding the challenge 10 Recommendations 11 Authors
Transaction Banking 3Research approachThe sample was drawn from ten different organisations,eight of which are financial institutions. In all, 14 differentreplacement programmes were considered.The information used to prepare this paper was gatheredfrom qualitative interviews with key IBM experts engagedwith our clients on core systems replacement programmes,in order to understand what the current best practice is forglobal platform replacement.While the majority of the programmes researched areglobal core banking solutions, we also interviewed IBMpeople who were engaged on platform replacementprogrammes with telecoms and industrial companies,to provide comparative perspectives.Summary of key findingsThe delivery of a global service is not a big driverfor core system replacement.Cost and risk management far outweigh revenuegrowth ambitions in the objectives for a global platformreplacement programme.Providing a single global solution to meet the needs ofan organisation’s global clients was a consideration only forone organisation we talked to that had attempted a globalplatform replacement programme.Replacement of global core systems takes a long timeand is very expensive.In spite of our aim to uncover and share best practice ondelivering these programmes more quickly and at lower cost,we found that significant underestimation is endemic, withprogrammes typically taking twice as long and costing twiceas much as originally expected.Our research suggests that organisations intending to replacetheir global core platforms should expect a programme thatwill require the following:•• Very large investment, measured in hundreds ofmillions of dollars.•• A significant amount of time, usually at least a decade.•• Very considerable management effort with sustainedboard sponsorship.Full organisational alignment with a programme iscritical to its chances of delivering successfully.The investment required for a global core platformreplacement means that it needs to be considered asa business problem, not an IT problem.The single most common characteristic across allsuccessful programmes is very strong business sponsorshipand a willingness to be patient and keep paying the bills.There is no dominant architecture, but a strongarchitectural approach is a prominent and enduringfeature of successful programmes.We identified no single architecture for a global systemreplacement that appears to offer a higher assurance ofsuccess, as well as no obvious vendor solution and nodominant implementation style.However, successful programmes implemented solutionsthat are closely aligned with the business case and madethem work through strong architectural governance.Outstanding challenges: our viewIn developing this paper, we have identified a set ofoutstanding challenges that the industry still needs toaddress in order to fully benefit from the investmentthat has to be made.There is a mismatch between time being taken andbusiness needs.Programmes that take at least a decade to complete donot appear to be a viable strategy for global core platformreplacement. This is because the organisation commissionsthe programme to address some fundamental issues whichcan’t wait that long.Architectures are replicating current design limitations.The solutions global organisations are building share someof the same characteristics as current solutions and could takea decade or more to replace in the future.
4 Global platform replacement: practice, issues and recommendationsThe needs of globalised customers are not beingaddressed.Constraints in the approaches taken and the solutions availablemean that global organisations are not tackling the needs oftheir customers who have commercial interests in the world’skey trading geographies.Difference, rather than synergy, is driving solution design.More focus is being put on managing functional differencesbetween geographies rather than exploiting synergies fromtheir similarities.We have concluded that without a major rethink ofprevailing practice and approach, there is a very realrisk that some major financial institutions and globalorganisations from other sectors will find global platformreplacement to be insurmountable and find themselvesunable to accomplish something centrally importantto their future success.DiscussionHow did organisations get here?The number of different ways organisations have arrivedat their current global platforms is as varied as the numberof organisations we have looked at.However, three patterns dominate the list, and everyorganisation in our sample fits into at least a combinationof the three:•• Acquisition-led consolidation: Organisations that haveaddressed global expansion through acquisition have gaineda new set of systems with each acquisition; in many casesthey have gained multiple new systems. Those that have beenaggressive about consolidation now have multiple instances ofa core solution, and those that have been less aggressive havethe added challenge of consolidation into their globalreplacement programme.•• Platform extension: Many organisations have entered newgeographies through the customisation and implementationof existing solutions. This may have been accomplished bythe enhancement of a single code base or by introducing newbranches of code.•• New solution: New solutions are implemented to supporta new geography. This approach is adopted for one of tworeasons: either because the cost and time required to amendexisting solutions is seen as prohibitive, or because a newgeography is seen as a potential pilot for a new solution.Irrespective of the ways organisations arrived at the systemlandscape they have today, each global core replacementprogramme will require a combination of migration andconsolidation.Drivers and inhibitorsThe delivery of a global service is not the key driverfor core system replacement.When asked why their organisation was embarking on aglobal core platform replacement programme, deliveringa platform that could provide global clients with a globalservice was identified as a driver for only one organisation.
Transaction Banking 5However, these three answers were given significantly moreoften than others:•• Reduction in the cost and risk associated with runninglegacy platforms. With many of the platforms in questionbeing over 30 years old, organisations are seeing accessingrelevant skills as a significant concern.•• Supporting the standardisation of business operations.Organisations see standardising operational platforms as anopportunity to decrease costs by reducing duplication withinoperations and ‘smartsourcing’ common activities.•• The provision of a platform that will support growth.Organisations embark on platform replacement to removeconstraints in current systems which inhibit the introductionof new products and entry into new markets.A smaller number also cited compliance, governance,consolidation and the ability to implement a step changein capability as reasons to replace their core systems.We suggest that the organisations may be resisting developingplatforms for global customers for these reasons:•• A judgement that the commercial benefits of providing acommon service to global customers will be outweighed byadditional costs incurred in complying with regional andcountry-specific laws and regulations.•• A belief that information consolidation to provide globalviews can be achieved without the implementation of acommon global platform.•• A perception that global customers are not demanding aglobal service and so they are not prepared to develop one,supported by a belief that continuing to use the currentbusiness model is a viable option.We think that these judgements may prove to be misplaceddue to the following reasons:•• As customers become more globally integrated, their needswill also globalise.•• Technology-enabled relationships between organisationsand their clients are already transforming what each expectsof the other.•• Innovation is now expected as the price of marketparticipation.Cost and timescalesReplacement of global core platforms takes a long timeand is very expensive.A driver for this research was a concern at ever-increasingestimates for the time these programmes will take. While weanticipated that it would be possible to identify best practicesthat could help organisations to deliver in a shorter timescale,the findings run counter to this, with successful programmestaking at least a decade.We suggest that this is down to a mismatch in expectationsbetween how complex a global platform replacement iswhen compared with platform extension or consolidationprogrammes with which an organisation may be more familiar.Extending the Platform FootprintExtending the use of an existing system to a greenfieldclient base in a new geography can be accomplished relativelyquickly, particularly when the products to be sold alreadyexist within the platform.By contrast, replacement programmes are more difficult.Solutions need to be configured to meet the needs of anexisting client base and product set. For an organisationthat has already rolled out to multiple countries, the costof meeting this need can be substantial.Consolidating onto a single platformWhere expansion has been accomplished through takeoverthe migration from numerous different solutions to a single,existing platform is a common strategy. In this scenario eachadditional country is regarded as a separate project andorganisations accept that a number of different countrieswill take a number of years.By contrast organisations expect replacement programmesto be more straightforward as the consolidation has alreadytaken place. This expectation can ignore the diversity ofbusiness rules can exist within a solution.
6 Global platform replacement: practice, issues and recommendationsReplacement programmes require a blend of the twoapproaches. The replacement of incumbent systems requiresglobal consolidation techniques to manage the businessrules, supported by roll out activity that is consistent withprogrammes to extend the platform footprint.OrganisationFull organisational alignment with a programme iscritical to its chances of delivering successfully.We found that the single most common factor acrossall successful programmes is strong business sponsorship.IT-initiated programmes, even with the best intentions, donot succeed in programmes of this scale. We observed that itis the drive for a consistent business model that results inthe ongoing support for a common IT system.Successful programmes, we found, align their rollout planswith the organisation to which the solution will be rolled out.If the organisation responsible for the transformation did nothave some authority over the countries where the solutionwas to be delivered, the programme was significantly moredifficult to deliver.•• Countries or branches may believe that their needswon’t be met by a global platform and push for their ownunique solution.•• Poorly-thought out chargeback mechanisms, where internalcross-charging for systems is not aligned with the systembenefits, can discourage adoption.•• Poorly-aligned executive objectives. For example, a businessunit accountable for its P&L may object to meeting costsof a new platform that it does not believe it needs.Getting a new platform accepted requires a significant amountof work from the delivery organisation and this can be madeimpossible if the whole organisation is not behind the change.Architecture and solution approachThere is no dominant architecture, but a strongarchitectural approach is a prominent feature ofsuccessful programmes.Solution approachWe found no solution that is more likely to result in asuccessful programme than any other.•• While package solutions are more common than bespokedevelopments, they do not appear to offer a greater chanceof success than a bespoke development.•• Of the package solutions available, no single vendorsolution appears to be dominant.•• No single implementation style appears to increasethe likelihood of delivery.PackagePackages reduce the development required, which isattractive to many organisations. However, they also comewith the following additional problems:•• Packages focus on the whole problem and require aconsiderable number of capabilities to be implementedin order to deliver a coherent solution.•• Packages introduce duplication with existing capabilitythat needs to be managed; for example, in relation tocustomer files.•• Integration can be significantly more complex than fora bespoke development.•• The change lifecycle can be considerably longer forrequirements which the package does not meet. Alternatively,the business might have to compromise on functionality.Despite these disadvantages, organisations continue toview a bespoke approach as carrying significant risk.VendorNo current vendor seems to have solved the issue of aglobal solution, either from a functional or non-functionalperspective.While progress is being made, it is difficult to identifya vendor package solution with the scalability required tosupport the largest organisations.
Transaction Banking 7Implementation styleThe choices of architecture for the solutions we researchedare shown in Figure 1 and Table 1 as four simple patterns.We found that the solution choice in banks was driven bythe way the organisation is structured and the will to pursuea specific course, rather than an analysis and understandingof the cost and benefits of a specific solution.For example, it was apparent that forcing a singleimplementation upon a firm organised on geographicpillars was not likely to work.Outside financial services, we found a trend whereorganisations attempt to implement a single global instance.While this is perhaps more difficult for a bank to achieve thanfirms in other industries, we expect that it is only a matterof time before a bank develops and adheres to thearchitectural patterns required.Two other interesting factors were also observed:•• Successful programmes focused on developing thearchitecture for the solution before the programme startedand backed this up with strong architectural governancethroughout.•• These programmes tended to replace old monolithic systemswith new, slightly less monolithic systems.While the newer systems were more modular and better atproviding services, in most cases their eventual replacementwill, in our view, require a programme of similar scale soonerthan a highly modular solution architecture.Easiest to implementHighest TCOHardest to implementLowest TCOConsolidationof regions onto asingle instanceSingle codebase rolled outacross regionsBase solutioncustomisedby regionUniquesolution in eachregionMost Solutions use acommon code baseLeast preferred Most preferredFigure 1: Solution options
8 Global platform replacement: practice, issues and recommendationsPattern Characteristics ImplicationsUnique solution for each region A common pattern within the platformswhich are being replaced. Often drivenby an acquisition expansion strategyso each country or region implementsits own specific solution. Regions andcountries have significant autonomyover change.No organisation has selected thisas an implementation pattern.Global services become difficultto deliver.Commonality driven by easeof implementation.Single source customised by region Another common pattern for platformsthat are being replaced. Origins inexpansion of an existing organisation.Same underlying application deployedbut region-specific customisationsmean that separate code bases aremaintained.Preferred by organisations that wantto leave some regional autonomy.Allows changes for one country to bedecoupled, but drives up maintenancecosts through duplicated effort.Multiple implementations of a singlecode baseRare in current solutions, but doesexist where an organisation has hada single-minded focus on maintainingone solution.Single code base is implemented inmultiple regions/countries with codebranched to support differences.Change process constrained by theneed to keep all in sync.Integration becomes morestraightforward, with a single source.Testing becomes increasingly complex.Single implementation We found no organisation startingfrom this position.A single instance of a code basesupporting regions and countriesaround the globe.No financial services organisationis planning on a single instanceimplementation, though we areseeing examples in other industries.Availability requirements of a global24/7 solution are difficult to meet andthe ability to manage the differentbusiness cycles, e.g. end-of-dayprocessing makes it complicated.Table 1: Solution options
Transaction Banking 9Understanding the challengeWe have concluded that without a major rethink ofprevailing practice and approach, there is a very realrisk that some major financial institutions will findglobal platform replacement to be an insurmoutableproblem and may find themselves unable to accomplishsomething centrally important to their future success.We found many factors preventing organisations from movingahead with global core replacement programmes:Cost and timescaleThe cost and timescale required to address global replacementsis a significant concern. While some banks are succeeding,there are more examples of failure which leads us to questionwhether replacement is a viable approach across the financialservices industry.For banks that don’t believe an incremental approach willaddress their needs, the challenges of a replacement may meanthey reach a stalemate. They have a set of things they believethey must do, but aren’t able to achieve.Ten years is a long time to wait to deliver business benefits.While benefits can be delivered incrementally, theseprogrammes are often set up to address fundamental issuesthat will not be fully solved until the programme is complete.Inability to handle scale of changeCurrent techniques for managing this level of change arepushing banks to implement solutions that have the samechange-inhibiting characteristics as their existing solutions.Instead, banks are implementing packages or core systemsthat provide a set of integrated capabilities that have to bedelivered as a set. The capability spike that this requiresstifles any ability to accelerate change.Local market complexityThe following other concerns are also deterring organisationsfrom replacement:•• Different accounting practices•• Consumer law and regulation•• Different regulatory frameworks•• Differences in market-side solutions, e.g., payment schemesWe believe that the tools to address these concerns existand are already understood across the industry, but we havefound no evidence that they are being properly used onreplacement programmes.While many existing packages support service-orientatedarchitectures, these are not underpinned by the granularityof components that allows a truly component-based approachto the problem.
10 Global platform replacement: practice, issues and recommendationsRecommendationsFrom the research undertaken, we are able to make a set ofrecommendations which should improve the likelihood ofsuccess with a global core platform replacement programme.1. Modernise rather than replaceIBM recommends a modernisation approach to core platformmigration, and our position is no different if the solution needsto be implemented globally. A modernisation approach thatexploits the best of what exists and prioritises change on theweakest components should do the following:•• Allow benefits to be delivered earlier.•• Mitigate the risk associated with waiting a long time forthe first implementation.•• Facilitate future change by increasing the modularityof the solution.•• Allow a bank to be more responsive to changes in themarket and adjust the course of the programme.2. If it is not possible to modernise and replacementis necessary, we recommend that organisations focuson getting these aspects right:Business caseEstablish a business case for the replacement programmethat will stand the test of time. If the business case won’tsurvive delays or significant budget increases, it won’t bestrong enough to outlive the project.Make sure that the business case is incremental and realistic.Confidence in delivering against initial promises will makesubsequent releases more viable.Build and sustain business support across the globeBusiness buy-in to the delivery plan and associateddisruption is required across all countries. When youhave defined a plan, you need to be able to stick to itwithout the fear of being derailed.Sustain consistent board sponsorshipBoard-level commitment is necessary for the investmentand the timescales.Architect for the future, not the pastArchitect the solution for the future: don’t rebuild the bankor firm of 1980. New solutions need to be significantly easierto adapt and expand than the solutions they are replacing.Identify which aspects of the solution must be global andchallenge what needs to be local. Drive out the benefits tothe solution and the delivery process that can be deliveredthrough synergy.Design for globalised customersMake sure that the programme considers the impact onglobal clients and doesn’t leave the solution at risk of a gapin capability against the competition. It could be a longtime before you get the opportunity to fill it.Build a global platformConsolidation programmes and rollout programmes tendto be more successful than replacement programmes. So,we suggest that you build a new platform and roll it out tobring other countries on board. The new platform should bearchitected as a global solution from the start and implementedand operational in at least one country before consolidationor rollout can start.
Transaction Banking 11AuthorsSimon GregsonExecutive IT ArchitectIBM Global Business Services, UK and Irelandsimon.email@example.com+44 (0)7739 876210Michael DavisonIndustry SolutionsIBM Financial Services Sector, UK and Irelandmichael.firstname.lastname@example.org +44 (0)7715 161155Simon Gregson is Executive IT Architect in the Bankingpractice of IBM Global Business Services, UK and Ireland.He has held the role of Chief Architect for two UK financialinstitutions and oversaw the technical transformation forEurope’s largest banking migration. He teaches ProgrammeArchitecture on the University of London’s Advanced Analysisand Design masters degree course. Within IBM, he specialisesin the technical leadership and management of IBM’s largestfinancial services engagements and leads the developmentof payments architecture and financial markets solutionsfor clients.Michael Davison leads the development of transaction bankingsolutions for IBM’s financial services sector, focused on coresystems replacement and complex programme management,on which he tutors for IBM in the UK and Europe. He hasworked on major change programmes in the UK’s leadingbanking groups over a 24-year period. He also leads IBM’sbusiness model innovation research programme with theCambridge Service Alliance, working in collaboration withCambridge University, BAE Systems, Pearson and Caterpillar.
© Copyright IBM Corporation 2013IBM United Kingdom Limited76 Upper GroundSouth BankLondonSE1 9PZProduced in the United KingdomJune 2013All Rights ReservedIBM, the IBM logo and ibm.com are trademarks or registered trademarksof International Business Machines Corporation in the United States, othercountries, or both. If these and other IBM trademarked terms are markedon their first occurrence in this information with a trademark symbol (® or™), these symbols indicate U.S. registered or common law trademarksowned by IBM at the time this information was published.Such trademarks may also be registered or common law trademarks inother countries. A current list of IBM trademarks is available on the Web at“Copyright and trademark information” at ibm.com/legal/copytrade.shtmlOther company, product and service names may be trademarks or servicemarks of others.References in this publication to IBM products and services do notimply that IBM intends to make them available in all countries in whichIBM operates.Please RecycleBKW03018-GBEN-00
... Simon Prunet, Michael ... Global Platform Replacement: Practice, Issues and Recommendations an IBM White Paper by Simon Gregson & Michael Davison
Below are the current and official "IBM Social Computing Guidelines," which we ... computing platform. IBM ... IBM is a global ...
Avaya provides solutions across multiple channels and devices for better customer ... Avaya Aura® Platform; ... To meet expanding global business needs, ...
KPMG is a global network of professional ... we are bringing together the global shipping industry to discuss key issues at KPMG's inaugural Global ...
McKinsey’s latest thinking on the issues that matter most in business and management. ... McKinsey Global Institute ... Featured Insights.
Official website for U.S. DEPARTMENT OF DEFENSE ... Security Situation in Europe Growing More Serious, Eucom Commander Says
MARKET TREND. THE BIG PICTURE. View IBD’s The Big Picture to get exclusive analysis on what’s happening in the market. STOCK MARKET TODAY VIDEO
... Leading provider of IT products and services for the global ... White Paper; Contact Us; News ... Fujitsu Laboratories Ltd. today announced that it ...
... from lecture notes to practice ... the company expanded into an online tutoring with a Q&A platform that ... "The best part of Course Hero's ...