Hard as it is to believe, the concept of return on investment(ROI) has been around for a while. Yes, its had different namessuchas business justificationbut even in the good old days companieswere reluctant to spend money without some idea of how theinvestment would be recouped. Today, though, the acronym ROI is onthe lips of every technology and business leader. But what is ROI?The term seems simple enough, yet those who work to determine thebenefits of a major project know there is nothing simple about it.Different companies have different ways of evaluating potentialinvestments, says Jack Tyniec, managing director of TCi Consulting& Research. Those criteria may not be the same for aninvestment like, for example, an acquisition of another company vs.the acquisition of a new technology or an improvement in a processinternally.

Bill Taylor, vice president of operations for AXA Re P&C, hasbeen in the insurance industry for over two decades, so he is usedto pushing through projects without the benefit of any ROImethodologies. He has seen the value of such a program, though, andunderstands why insurers are buying into such methods.

It depends on what hat youre wearing as to what kind of value yousee in these things, he says. As these methodologies and models arerefined, it does make business users think about things thatperhaps they hadnt thought of before.

The metrics for ROI change from company to company, asserts JulieDorey, vice president insurance and financial services forconsulting group Xerox Global Services. There is more data forcompanies to work with today, she says. Its not just a form to befilled out with a purchase requisition. Its a complicated processthat not only includes facts and figures, but work processimprovements in many cases. The metrics today are changing as moredata is available and as we are measuring different things.

How to Do It
In 2001, the insurance and financial services company AXA developeda methodology for ROI that calculated and required certain tasks tobe completed if projects are of a certain size, according toTaylor. AXA has four levels of investment categories: low (aproject costing $25,000 or less); low to medium (less than$200,000); medium to high (less than $2 million); and high (morethan $2 million). The first level does not require calculating ROI,but rather an executive summary of the project. The second levelinvolves a simplified business plan, which basically is ascaled-down version of what takes place at the third and fourthlevels. The top two levels require what AXA calls Risk AdjustedValuation of Investment (RAVI), according to Taylor.

This ultimately feeds into what [AXA] calls a business-caseapplication, says Taylor. It is a series of sophisticatedspreadsheets that drive a process. It makes you think about thingsyou probably wouldnt have thought about [previously] or put down[on paper].

The RAVI system begins with an outline of the project definition,according to Taylor. Within that outline, subjects are addressedsuch as why change is needed, the objectives of change, the scopeof change, whether or not there are alternative solutions, andlimits or constraints to the change. [The outline] also asks forother considerations, says Taylor, such as what kind of revenue theproject will generate. They want to know specifically if there isan increase by premium, a decrease in loss ratio, a decrease ingeneral expense, or if it is mixed. They also want to know if its aregulatory-related project or compliance-related initiative.

Some Things Dont Pay
There is nothing that destroys ROI more than regulatory changes or,in some cases, strategic changes. Taylor says such factors as finesfor noncompliance can be factored into ROI, but regulatory andstrategic issues often go beyond ROI. Chris Haines, manager oftechnology operations for the Buckeye Insurance Group, says, I cantput a monetary value on stat reporting. Now, if were getting fined,I guess we can.

The compliance area falls under the risk column for insurers, saysDorey. Sometimes you cant measure ROI because the risk ofnoncompliance is so high its not pertinent to measure. If you dontconform to regulatory compliance, theres no point in doing an ROIon it, she says.

She points to the Sarbanes-Oxley reforms that focus on corporateaccountability as a good example. Everybody is rushing to meetdeadlines there, so Id say [insurers] still are looking todetermine an ROI for pieces of that [legislation], especially thetechnology piece, she says. Yes, insurers are seeking ROI, butintuitively they know they have to comply.
Technology has that effect on insurers, she believes. Whentechnology becomes involved, you see the ROI come up right away. If[a project is] more process innovation or behavioral change, youllsee fewer ROI requirements.

Projects that might save a carrier money or lower expensessometimes have to be delayed to meet these issues.

[Compliance] is an expense. You have to bear it, and it may affectthe amount of money you have available for discretionary projectsbecause its a have to do kind of issue, says Tyniec. In the areasof compliance, companies have had to make a lot of changes, havehad to invest a lot of money in both business processes andtechnology to make sure they are writing quality business andprotecting the customer. These require substantial investments, andthey arent typically justified on the basis of any kind of ROI.They are issues where you just have to do it.

Keeping Up With the Joneses
Another worrisome area for those hoping for a good ROI involves thecompetitive pressures that insurers deal with. A company isoffering a particular kind of capability on its Web site or accessto its portal, and you need to match that, says Tyniec. Its not aquestion of do I or dont I do it, because if you did thecalculations, you would find yourself losing revenue if you didntimplement some of these things. Thats a qualitative assessment.Some companies will carry that out to a quantitative analysis, butin many cases, senior marketing people and senior executives of thecompany simply are saying, We dont have a choice. Thats thedirection the industry is going. We need to at least match thecapabilities our competition or peer companies are providing, andtherefore we need to do [the project]. You could decide, if youwant your company to go to hell in a handbasket, to do without[certain capabilities], but typically companies are a lot smarterthan that.

As an example, Dorey mentions currently working with a client on aproject that is strictly strategic in nature. You cant put an ROIaround having a strategy, she says. The VP doesnt want to have anyconversation from anyone on the team [about ROI]. Thats more of ararity.

Strategic Information
Devising a corporate strategy may go beyond the justification of astraightforward financial return, but AXA believes its ownstrategic information has to be contained in parts of its RAVIreport. Such data is broken down by client information,distribution information, employee information, technologyinformation, and the manufacturing and processes, according toTaylor. You go through a whole series of questions, he says. In thecase of the client assessment, you have to list various clientvalues as well as client risks. You come up with a score and risksfor all five parameters.

AXA then examines the tangible benefits of the project. These arebroken down into three main componentseffort savings, cost savings,and the business benefits, according to Taylor. Listing these areascan be a daunting task, but Taylor says the company has identifiedcertain areas the ROI team can select from. For instance, in theeffort-saving area, there are [multiple] major areas you can choosefrom to put in actual dollar values for year one through year five,he says. One of the examples under effort savings would be thereduction of task-processing costor the improvement ofproductivity. Numbers for reduction of efforts, current averageefforts, and annual occurrences would be included in this area.There are eight different areas you can choose from in theeffort-savings part, says Taylor. [For] a typical project youdprobably choose only one or two things in each of the threecomponents, but these quantify tangible benefits.

The financial information doesnt come into play until the fourtharea of the report, says Taylor. This involves financial data fromthe enterprise or entity as far as revenues, salaries, andexpenses. There is some interpolation as to the cost and savingsfor the project at the financial level, he says.
The final piece of the AXA puzzle is called Advice for Tracking ofRisks and Benefits. You have to demonstrate how you are going toreduce risks of things youve identified in the strategicinformation gathering, says Taylor.

Lets All Do It
Many smaller companies would drown in work if they attempted aproduction similar to the AXA effort, Buckeyes Haines says. Hiscompany is not alone in this regard. A lot of companies in our peergroup that we interact with arent doing [complicated ROI formulas],eitheras far as a drawn-out mathematical theory that is used tocome up with a number. He claims smaller insurers instead have tofocus on hard deliverables and milestones to justify expenses. Inthe last five years, weve been able to grow substantially as wellas become more profitable with a decrease in staff, he says. Wegauge things by those types of results rather than an actualmonetary value.

Buckeye is getting ready to start a new project that will involveonline rating, quoting, and endorsing for the independent agenciesthat sell the insurers property/casualty products. The things welook at as ROI for us, Haines says, are whether we will be able toeliminate X number of staff as well as grow by a certain amountwithout reaching a breaking point.

While he knows he could quantify those working hours saved by newtechnology or processes, he believes some of those figures fallinto a gray area. As much as I would want to show that in an ROIfor technology, somebody else might want to show [those savings]for something else.

Hard vs. Soft
Some benefits are determined more easily through the ROI processthan others. Finding the hard-dollar savings from a technologyproject is one of them. But technology itself has helped insurersfind soft-dollar savings, too. Stretching those numbers to make fora more attractive ROI can be a dangerous practice, though. I thinkpeople are tougher now, says Dorey about scrutinizing financialprojections. She believes managers are going to challenge numbersthat cant be easily substantiated.

They are more personally accountable, especially in this industry,she says. The insurance industry is based on peoples trust levels.So on one side I believe people are tougher and more careful aboutwhat they put in their ROI, but on the other side there is a lotmore data available to help bridge things into the equation thatwould have been considered soft dollars in the past.

Technology has enabled insurers to capture information aroundvirtually everything within the company, Dorey believes. You cancreate an ROI today that includes information that in the past youcould never include, she says. Its all contained in databases thatare accessible and real time. The content management and documentmanagement systems that are out there are enabling people to createROI for projects or work processes. Yes, people are able to createan ROI that includes things they never could include in the past,but on the other side they are very cautious about what they put inthere.

Numbers are funny, Taylor points out. You can make them fit.
Insurers have the ability today to turn qualitative analysis intoquantitative numbers, Tyniec remarks. On a technology project, forexample, say you are looking to implement something that is goingto allow you to sell more products, he says. You need to crank inwhat the benefits are of a revenue expansion. On a cost-oriented ITproject, its difficult to translate revenue impact into a projectscash flow based upon expenses because the revenue piece may havelittle impact on the expense and what you are looking at is a largeexpense with no expense savings impact.

Tyniec believes insurers need to focus on the very top line of thecompany where a project is going to have an impact on companyrevenue and ultimately find its way into the bottom line andimprove earnings. If the project is successful and product salesincrease, the expense required to process and administer thoseproducts likely will increase. But thats a good thing, says Tyniec.Companies thinking creatively and analytically about this recognizethere needs to be some way of acknowledging and including thesesoft dollars or nonquantifiable aspects into an evaluation of an ITprojects overall viability.

Making It Work
Taylor is impressed with the sophistication of the AXA system. Itallows the company to compare return over risk for each of the fivestrategic values it has identified and come up with an overallscore. AXA believes in sharing information rather than reinventingthe wheel for each project, so project managers can reuseinformation already compiled for a separate project. Theres alsothe issue of reuse of business technology, he says. AXA has its ownGlobal Information Technology Organization (GITO) that is astandards and guidelines body for all the AXA entities. Bypublishing the RAVI information within AXA, the GITO is alerted tohow the business side is going to use new technology and can offersuggestions how the technology can be used in other ways that willfurther reduce expenses. So theres a method to the madness as faras being an ROI methodology, he says. It helps a large organizationpolice itself, as well.

While appreciating the methodology, Taylor says he is glad the datawarehouse project he completed in 2000 was done without the benefitof a detailed ROI. When we did our data warehouse initiative, wedidnt have any formalized process, he says. We were a relativelysmall entity within the confines of this large organization. If wehad done this [data warehouse] project a year later, I would havebeen under more rigid guidelines as far as what needed to be donewithin the AXA organization.

Flying without the ROI net really didnt bother Taylor, though. As aperson doing this for a number of years, Im not sure how valuableit is, he says. I was pretty self-confident in what I felt I neededto do and how to attack it and didnt want something of this natureto get in the way. On the flip side, if I were the controller, Iwould want this type of [ROI] project done.

IT for ITs Sake
Insurers have to look at multiple savings when an ROI is conductedfor an IT project that has more of a business impact, Tyniecbelieves. IT often proposes projects that have to do only with ITand have only IT benefits, he says. These are relatively difficultto justify unless they actually reduce costssuch as a newtechnology that is more efficient and cheaper.

He adds ROI comes into its own, though, when the project involvestechnology that has more of a business impact and affects theeffectiveness and expense structure of the business that is beingsupported. In most cases, when you do an IT project, there ought tobe a business benefit, and that business benefit ought to bequantifiable, says Tyniec. The easiest case is where you propose anew system that has improved processes and technology and is goingto reduce costsreduce costs in the business area.

Tyniec runs across businesses that dont always consider thebusiness impact of a particular technology, he says. The analysisof the ROI is confined to what it is doing within IT, and thats notthe end of the story, he says. The impact, to the extent that itexists in the business community, really can have a magnitude thatdwarfs the benefits inside IT. If you can quantify it that way, youreally have a powerful tool.

And that tool will make the next project down the line that mucheasier to get accomplished.