We want to get a project justified, but the ROI model is not helping us since this is a replacement of a very old, antiquated and out of date system. There will be no increase in revenue, no process change, just a replacement of technology. Is there any way to justify this with an ROI.

The business is on-board with the project, the sponsors are almost on-board. How do we get the financial people on board?

4 Answers 4


A couple of anecdotes might help stimulate thinking about how to pitch this to the financial guys...

  1. A large financial client of ours had some older servers running in the back area. The applications were working fine, and weren't in any need of updating, until the 2007 change in daylight savings time. It turned out that some of the older servers were NT 3.51 which was no longer supported by Microsoft. As some of the applications could not be migrated to newer operating systems, this financial client had to pay Microsoft $350,000 to write a patch for them to support the new Daylight Savings Time rules.

  2. At my previous employer, they sell financial software used by accountants and actuaries. One of the applications was still being written in VB4 and used VBXs (16 bit ActiveX controls). When Vista came out, this application could not run on Vista, and we only found out about the problem when customers started calling the help desk. It was a year before we had a version ported to .NET so it could run on Vista and later operating systems. We suspect that VB6 applications (VB6 apps make up 60% of the revenue of this branch office) will have a similar problem when the next version of Windows comes out.

  3. At a company I worked at about a decade ago, they had licensed the Altavista search engine. Somehow, in various moves, the installation media got lost. When they were building a replacement server, they approached the new owners of Altavisa who said they no longer licensed it at any price, so get lost.

My suspicion is that you'll have to pitch it in terms of risk: what will it cost when something out of your control breaks what currently works?

  • 1
    Thanks. Yes, risk is the best we can do. I was trying to get from the community a way of measuring the risk in dollas so that I can put it inside the ROI. Maybe trying to fit the ROI model is what is causing so much issues. Thanks again for your stories, I think these things happen often.
    – Geo
    Feb 12, 2011 at 1:44
  • Risk of non-compliance with new regulations can be quite high, too. That is industry-specific, however.
    – SBWorks
    Feb 16, 2011 at 8:43

If the old system works reliably and is valuable to its users, if maintenance costs will not be reduced by introducing a new system, if the new functionality will not have any financial impact, if there are no savings for moving the software from an old platform to a new platform, if there are no regulatory or non-compliance risk with the old system, why do you want to replace it in the first place?

Why do you say the system is out of date? Just because it is written in an old language? If there is no maintenance risk, then why replace it?

Will the system run faster or will be able to run on less expensive hardware if you rewrite it? If yes, maybe you have a case.

Remember that rewriting the system will bring you back to square one on the learning curve, you will replace a reliable and debugged system with one that may have many bugs and may not implement correctly the current functionality.

  • Great point of view. One of the biggest mistakes (at least on software business) it to write the hole system or application from scratch. This is one of things you should never do by Joel Spolsky. Of course that this might not be the case of Geo, but should be aware of it anyway.
    – Johnny
    Feb 14, 2011 at 12:47

The first place that I'd look for savings is in the cost of labor to maintain the old system. Is the labor pool for the old technology shrinking? If so, people in that pool are probably paid a premium. Can you save enough on labor in a short number of years to justify the initial cost?

  • Great question Bill. The answer is no. The company employees all the people with the expertise to maintain the application. The only cost that we are able to justify, is if the application fails today, we will have to do all the things manually which will cost lots of time and delays to the business. Does that answer your question?
    – Geo
    Feb 11, 2011 at 23:11
  • @Geo: I meant that to be the questions you should be asking yourself to find some ROI in updating an old system. Feb 12, 2011 at 22:03

ROI is vague as one may deliberately use it to compare almost anything: 1. Revenue/capex? 2. PBIT/opex? 3. CPU/cost? 4. you name it.

Another approach is by using DROI (Discounted Return on Investment). It is simply NPV over capex. Because NPV is used, DROI recognizes time value of money and can be used to rank e.g. projects. However, this also has drawback as one has to really sure these projects are comparable. Also there are rule of thumbs on how big is "big" and how small is "small"and they differ among industries (and projects). For example, offshore deep water oilfield development in concession fiscal regime with more than XXXMMstb reserve needs to have DROI ranging from 0.2-0.5 to be deemed viable (note how I clearly specify the explanation).

Regards, Ady

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.