Some of you know that we (Summit Business Media, TechDecisions, and CastleBay Consulting) recently completed aseries of webinars as part of our 2013 Shop Talk Communityinitiative.  These webinars followed the Shop Talk bookstructure and focused in turn on the vendor software market,software selection, and core system implementationprojects.  This was an opportunity to update our thinkingon these relatively dynamic areas.  Here is the mainnarrative which emerged from this exercise.

Not Your Father's Software Market

The core systems software market has changed radically and forthe better in recent years.  Indeed there has never been abetter time for a carrier to acquire third partysoftware.  There are several reasons for thisassertion. 

First, vendors are writing better and more completesoftware.  The functional footprint of a policyadministration system has always included product definition,rating, policy maintenance, and print functions.  ModernPAS often include significant functional additions such as portals,for agents and consumers, BI capabilities, and other majorenhancements. 

While the base functional footprint has grown there also isrecognition that a "package system" will not meet the specificneeds of any potential customer and that bridging the delta betweenthe base functionality and the carrier's requirements is criticalto the success of the implementation. 

This fairly obvious fact led to a profound shift in how vendorsthought about and delivered software and hence to the developmentof configuration toolkits which can be used to close that defineddelta more quickly and at lower cost than everbefore.   Further, most core systems now supportboth personal and commercial lines solutions; as recently as sevenor eight years ago this was not the case. The hoped for result is ahighly customized system at "packaged software"prices. 

Second, the carrier now has more acquisition and implementationoptions.  Vendors commonly offer in-house licenses, ASPmodels, and are increasingly moving toward the provision of SaaSand cloud solutions.  Also, the days when "rip andreplace" was the only implementation option are gone as vendorsprovide "best of both"  offerings that blur the linesbetween best-of-breed and suite solutions and allow carriersgreater freedom in deciding how and when to takesoftware. 

Third, vendors generally have moved to a more agile style ofimplementation methodology which increases the likelihood of betterimplementation outcomes by encouraging early viewing and feedbackcycles as the software evolves.

Success Needle Hasn't Moved

Given the above, you would expect that the industry'simplementation track record would have improvedsignificantly.  It hasn't; or if it has it's not asobvious as it should be.   A recent study byconsulting firm PwC concluded that: "Despite heavy investment inpolicy administration system modernization projects, less thanone-third succeed, half are 'challenged,' and one fifth of themfail." 

PwC defined the folllowing:

• Succeeded: The project was completed on time, within budget,and met all original benefit requirements.
• Challenged: The project made it to the deadline, butexperienced cost/schedule overruns and was unable to fulfill allthe original benefit requirements.
• Failed: The project was abandoned or cancelled due to projectbeing unable to meet cost, schedule, or customer expectations.

One can take issue with the definitions provided here and thereis an argument that  some of the projects that end up inthe "challenged" category should be moved to "succeeded," but thereprobably are projects in the same category that are closer to"failed."

The key point is that these types of studies, which despitetheir intermittent appearance and inconsistencies across publishingentities, reflect broadly similar results which have not changedmuch over the past 15 years. 

So, the significant improvement in development and delivery onthe part of the vendor community has failed to move the needle interms of overall industry success in core modernizationprojects.  Why?

How Do You Get to Carnegie Hall?

The single biggest problem faced by a carrier in undertaking acore systems legacy modernization project is the fact that theyundertake this kind of complex, enterprise-wide initiative onceevery 15 to 30 years.   

An old joke comes to mind: "Question:  How do you getto Carnegie Hall?  Answer:  Practice." Suddenly, the carrier is supposed to be capable of pulling off aCarnegie Hall performance having spent years playing "ChopSticks."  So while the vendors, who practice for CarnegieHall on a constant basis, are much more capable, their clients byand large are not.  Fifteen years of maintenance, bugfixes, regulatory changes and rate increases does not prepare acarrier for the rigors of core legacy replacement. This fact isreflected in some important ways. 

• Wrong- headed thinking.  Somecarriers still feel that having contracted with a vendor, thatvendor will run the project and somehow make the project asuccess.  Carriers must recognize that the vendor isresponsible for only a subset of the project, which does notinclude such crucial activities as defining requirements, testingand validating the software, planning and executing dataconversions, training staff on the new system, planning the rolloutof the software, and planning and organizing the overall project,including coordinating with the vendor.  In a recentpublication Blueprint for Success: Core System DecisionsSMA Partner Karen Furtado states: "An examination of the totaleffort involved in a core system implementation shows thatapproximately 30 percent of the time expended involves the actualcore system software. The remaining 70 percent of the effort isspent developing requirements, dealing with systems integration,managing the data, and testing and deploying the system." This 70 percent is almost exclusively the responsibility of thecarrier.
• Carriers lack expertise. Most insurers lackdiscipline and the infrastructure necessary to successfully executea core system project.    Furtado again:"Insurers need to assess their implementation readiness … A project management discipline is mandatory—one that givestransparency to the project … Experienced project managers …. thathave experience with a number of software implementations iscritical. Another crucial area …. business analyst resources thathave expertise in quality assurance and testingmethodologies…"  And of course these key resources are thevery people that every project wants, and they are also the peoplewho keep the business operations running.   Manycarriers mistake project administrators for project leaders andfail to grasp the important fact that if the project leader doesnot viscerally understand the goals of the project they may driveit to the wrong destination.  Similarly those key businessanalyst resources may know the business but they may not know howto develop and execute a formal test plan.
• Still lacking. Furtado again: "Experiencedproject managers are key resources. The knowledge of those thathave experience with a number of software implementations iscritical. Another crucial area for most insurers is the staffing ofimplementation projects. Key business analyst resources that haveexpertise in quality assurance and testing methodologies areneeded".  And of course these "key resources" are the verypeople that every project wants, and they are also the people whokeep the business operations running. 
• Think differently. Carriers need to start tothink about these transformational projects in a differentway.   The real objective of most core systemreplacements is to create a new, flexible platform which cansupport and even facilitate the growth of the carrier for years tocome, and yet our usual criteria for success is a list of businessfunction requirements.  Another aspect of this mindsetissue came out in our webinar on implementations when a colleagueof mine, Ralph Vagnoni, stated that carriers need to realize:"Budgets and timelines are based on estimates and assumptions, notfacts."  The implication being, for example, if theassumptions change so will the timeline and the budget. Rational as this statement is, it raised eyebrows andvoices.  Carriers, it seems, still want these hugeundertakings to be defined as delivering a list of identifiedfunctions on an unchanging timeline for a fixed budget. Hardly a recipe for success.  I am not suggesting we throwdiscipline out the window but even the best-run projects grow andchange for good reason and it seems that rigid thinking issometimes mistaken for disciplined thinking. 

Carriers face real challenges born of the Carnegie Hallprinciple, in executing core system projects, but the ongoingdismal results of surveys like the PwC report cited above may besimply how we look at and think about these projects. Partly it seems the way we define these projects makes itimpossible for many of them to succeed.