Buy or build? Its been an important question debated in IT shopsfor as long as there have been IT shops. But software vendorspatiently have chipped away at the builders of the IT world to thepoint that build is rarely the best option anymore. It used to be aquestion of, Should I buy or should I build? says John Flynn,senior vice president of META Group. Were past that question today.Now its, What do I buy and for what reason?

A third question could be: Whom do I buy from? For mostinsurance carriers the answer appears to be someone with atried-and-true product. As the insurance software market getscrowded, it is hard to find a product that is both exceptional andpersuasive, yet that is what insurers are asking of newcomers tothe market.

The insurance industry is all about managing risk, says KevinDennis, director of enterprise solutions for the consultantInteractive Business Systems, and jumping on a new vendor orsolution is riskier than sticking with a proven commodity. There isa tendency in any risk-averse industry such as insurance not topick the new guy, he says. Firms that have produced software for along timeCSC, Appliedare very good. They are reputable, you knowwhat youre getting, and they can give you a long list ofreferences. To break into the market you would have to have [aproduct] thats unique and compelling. Thats getting really hard todo these days.

Still, there are times insurers must face up to a difficultsituation and dig in with their own set of tools. There may be asolution to buy for every insurance problem, but getting it to fitwithin a carriers sometimes antiquated legacy system makes buying asolution a waste of money. Ed Leveille, CIO and vice presidentsystems at Providence Washington Insurance Companies, says hiscompany prefers to buy, like most other companies today. Sometimes,though, carriers cant find the right tools and are forced into adifficult position of building a system. We couldnt find a truepolicy administration system that fit our needs, he says. We sawthat anything we did buy we were going to have to do a lot ofcustomization to the front end anyway, so we just said lets goahead and build it.

New Guys
More and more products are being launched into the insurancetechnology market, and while the new and unknown scares somecarriers, Leveille says it doesnt bother him. In fact, he recentlylicensed two products for a Providence Washington start-up companyfrom a new vendor that hasnt even completed the product yet. Ihappen to know the principals of the company from way back, and Iknow their strengths and weaknesses, he says. Im very muchconvinced its going to be an excellent product.

Typically, when selecting a vendor, Leveille looks around themarket himself and consults research firms to give him a list ofthe key players in the market. Even after researching a company andits customer base, Leveille indicates it still comes down to a gutfeeling. Whats your gut really say about these people? he asks. Howdo these people deal with you? Is it more as a customer or more asa business partner? You can pick that up quickly in yourconversations.

Leveille has made deals with the top vendors in the market, hesays, and gambled with companies that dont have very many installedpieces. The proof of the pudding comes after the contract issigned, and you have the installation, and you give them the firstproject to assist you with. Whats the service like? How do theyrespond to you? If youve got a major production problem, do theygive you good service? I cant say all my vendors are like that, butthe ones that get most of my business are like that. They give meservice, and they treat me like a partner.

Nicholas Wilkoff, a senior analyst with Forrester Research,believes it is getting more difficult for new vendors to enter themarket. Companies are looking to work with vendors they havefamiliarity with and to standardize on specific platforms, he says.Companies are taking a look at the technology theyve invested inacross the enterprise and at the lower-level infrastructuretechnology to standardize platforms from IBM, Microsoft, or Oracle.This focus on standardization is hurting the best-of-breed vendors,he asserts, and forcing consolidations among vendors across thetechnology market.

Flynn points out it is a buyers market, as well. When you go outand look at the critical mass of solutions that are available, itsmore than a handful, he says. Now, in some segments, you see 30,40, 50, 70 providers. Such a crowded field is beneficial tocarriers. The new entrants are stirring the potwaking up some ofthe older sleeping giants and helping them recognize the need toinnovate and grow and replace their product as opposed to justmaintaining it.

Taking somewhat of an opposite view, James Fridenberg, vicepresident, policy management applications, with the FarmersInsurance Group, believes new vendors do not face an uphill battle.New vendors in the insurance software application space often havenew approaches that offer a cost-effective alternative to thetraditional players, he says. As an insurance company, we need tocontinue to look at all the options that are available to us.

Farmers evaluates products and solutions based on a number ofcriteria, Fridenberg says, such as integration with the carriersexisting software and hardware platforms, return on investment,installed base, compatibility with the Farmers business model,vendor reputation, staff training needs, TCO, and end-userevaluation. The new insurance solution providers will provethemselves by scoring well in these categories, he says.

Ralph Carlile, Grange Insurance Groups director of IT, assertsit is difficult for new vendors that offer only run-of-the-milltypes of applications. But if vendors can find a niche,particularly with a business-critical application, even newcomershave a chance to succeed. In the insurance industry, theres a hugeneed for a richer product set related to integrated policy, claims,and billing processes, he says. Right now, there are less than ahandful of companies that do that in the client/server world, andmany of them do not offer a configurable system. Thats why you seesuch a high level of proprietary software in the business-criticalfunctionality related to the vertical market.

The fact that these applications are business critical is onereason insurers have been reluctant to embrace new vendors. Theyare just not willing to invest in something unless it has a trackrecord, says Carlile. You are looking at something that is a corebusiness applicationsomething that your revenue stream dependson.

Insurers are unlikely to invest in one of these coreapplications without some firm assurances it will work. Breaking inwith a smaller carrier isnt the direction to go, either. Smallercompanies look to the larger ones, Carlile says. If the largercompanies have not implemented it, the smaller companies arentlikely to use it.

To the Core
Leveille still believes it is easier to find most systems on themarket and dismisses talk that a particular system can give aninsurer a competitive advantage. When I want a billing system or aclaims system, I plan to purchase it, he says. Claims is claims isclaims to me and to our business people. We think there are goodsystems out there to purchase and just plug in.
Forresters Wilkoff says just like much of todays technology, ITdepartments need to be flexible, as well. I dont think itsnecessarily a choice of whether you build or you buy, he asserts.In many cases, the line has become blurred, and there is the optionof assembly and building applications off of existinginfrastructure and technology.

Flynn notes clients of META Group still raise the build vs. buyquestion. They tend to put it in the context of a specificfunctional needsuch as rating engines, policy administrationsystems, or claims applications, he says. If you just think aboutthose three areas, they represent a fairly large percentage of theoperational function an insurance provider has to engage in.

Still, Flynn believes the ability to buy a best-of-breedsolution that supports the three functions above will be morebeneficial in the long run for carriers. Buying a tool allows themto deliver greater value to their user base so [the carrier] canfocus on the hard stuff, such as workflow enablement, as opposed towriting the functional code to build [a system].

An Easy Choice
Any consumer of IT services would much rather buy services, Flynnbelieves, whether that means a physical product or people, if thepurchaser believes the solution will deliver greater value to thecompany and diminish risk. The biggest reasons for this are thedepth and breadth of the software available and the state ofoutsourcing in the insurance market today. What youve seen over thelast five years is a greater emphasis on providers recognizingthere is opportunity in the insurance marketplace, he says. So yousee products emerging with fairly significant industryunderstanding and depth.

The big challenge for vendors over the years has been how dothey enter a marketplace with a product thats attractive enough tothe buyer community while, at the same time, producing a profit forthem. For the longest time, providers of software solutions haveapproached the [insurance] industry cautiously in large partbecause of how well they understood the [insurance] industry andhow much penetration they could get with the product, says Flynn.As they recognized that greater industry knowledge and depth getsthem the robustness and functionality [carriers seek], they can getbetter penetration.

Still, Wilkoff adds, some projects get carried too far, and thecost for the insurer gets out of hand. What carriers are doing isisolating the projects or breaking them down into more specificprojects, he says. If you look at something like compliance, forexample, when you get into the technology, it can be broken downinto a number of different products that can be used by aninsurance company to drive toward compliance, so [the carrier]focuses on electronic records management as a technology they wantto make a selection around.

Standards Are Important
While standards such as ACORD XML still are emerging, Flynnmaintains enough customers are demanding compliance that vendorshave little choice but to jump on board. Standards are nowhere nearcomplete, and by no means are they accepted 100 percent byeverybody, but it has reached enough critical mass where itsinfluencing the marketplace, he says. People are saying, Is yourproduct XML compliant? If its not, I wont choose it. Thats a greatthing.

Leveille also is a strong believer vendors need to react to thestandards movement. Ive gambled my company on it, he says. I think[standards] are slow coming, but every month or so I read anotherheadline where another major vendor in the insurance space hasdecided to go that route. There will be different variations andversions as we go forward, but the specifications and standardsthat are playing out in this whole XML thing are here to stay.

Dennis of Interactive Business Systems believes the standardsfight is a good one, adding the ACORD XML standards are a huge stepin the right direction for [insurers] actually to transmitinformation. But even though a standard is available, it doesntmean the problem is solved just yet. What everyone is strugglingwith right now is how do we agree on a standard to tradeinformation, he adds. Everyone wants [software] to come out withsomething that conforms to those ACORD standards. The problem isthose standards are extremely flexible. Just because a productconforms with a standard doesnt mean it is going to be possible totrade information.

Fridenberg agrees standards such as ACORD XML are veryimportant. It makes it easier to integrate third-party productsinto our applications, and interfacing with external vendors andservices becomes less challenging, he says.

Vendors have to be standards compliant, Carlile believes,particularly with the ACORD XML standards. Insurers have to havethe ability to integrate systems quickly if they are going to buysoftware off the shelf, he says. The XML standards probably are ourbest hope right now of being able to accomplish that [integration].As systems become more XML compliant, it would be nice if they wereACORD XML compliant as well if they are going to deal with theinsurance industry.

Still Scared
If there is one word that makes insurers leery about buying a newsoftware package, it is customization. Thats what I use as mydecision point, says Leveille. How much customization am I going tohave to do to the package to accommodate my business users? If itsa lot, then I dont want to buy a package because typically if I buya package I have to pay a maintenance fee and support servicesannually at like 15 percent to 18 percent of the license feethatswhat the markets pretty much at now. If Ive done a lot ofcustomizations, I cant do the upgrades anyway. It kind of defeatsthe purpose, I think, if youve bought a package and greatlymodified it.

Still, customization is done all the time. Leveille believes itis done for one of two reasons. Either the package was not a verygood fit to begin withit might have come close to what you wereshooting for but not near it, he says, or the users want the systemto fit into their current structure and are unwilling to changethat structure to fit the new system.

That is why many insurers are turning to portfolio management totake advantage of software they already own but may not be fullyexploiting. A lot of that is driven by acquisitions, says Dennis.Insurers arent really sure about what theyve got and what exactlythe software not only is doing, but what it can do. Some of thesefirms want to reach an economy of scale by putting policies on onesystem, and to reach that economy of scale, they have to figure outwhat is the best of breed they have and what is the easiest tosupport and how it integrates with the tools they are providing totheir constituency.

The Right Decision
The decision to buy a system today comes down to recognition by theCIO of what the company is trying to do with technology, saysFlynn. Youve got CIOs or business unit executives recognizing theindustry they are in is insurance, not technology, he asserts. Whatthey can be great at is insurance. They can be pretty good atdeveloping software, but at what cost? Is [building software] wherethey really want to sink their capital and manage their risk?

The mystery has been taken out of insurance software, as well.The [software] industry recognizes insurance is a difficultfinancial product. However, it is not so unique that commerciallyviable solutions cant be constructed around it, says Flynn. For thelongest time, it was believed insurance products or systems were sodifficult and obscure you couldnt create technology on the outsideas a commercial product to service it. That myth has been brokendown. There still are aspects of the industry that actually havethat challenge and probably always will, but its not the criticalmass of what is happening in the industry anymore.

Nevertheless, says Fridenberg, many of the off-the-shelf policyprocessing systems or claims administration systems on the marketare more suitable for small to medium-size insurance companies.Smaller companies typically can adjust and adapt their businessproducts and processes around a software product with littleimpact, he says. For a large multiline carrier such as FarmersInsurance, which operates in 42 states and offers many insuranceproducts in all lines of business, off-the-shelf products typicallydo not provide the level of robust functionality and performancecapability we require. For the most part, we have found buildingcustom systems provides us the best value proposition in cost,functionality, and flexibility. However, we still look and evaluateother products all the time as you always have to be surveying thelandscape.

The Final Answer
As insurance carriers learn to better manage their portfolio oftechnology investments, Flynn believes the industry will seegreater levels of judgment on business cases that proposetechnology. The questions will continue to be asked, he says. Arewe building this, or are we buying it? If we are building it, why?Dont avoid the questions, defend [the answers] rigorously. Thosewho build probably will have a very justified reason for doing it.Its just not so easy to build it yourself anymore. Theres too muchchoice out there.