IT-projects are expensive and have a tendency to fail or to become more expensive. That is a persistent image. And according to long-standing research also true in about 40% of the cases. A growing number of board members are demanding a proper substantiation of the investment in new IT. That sounds like good news. Yet too often we hear from board members that in their experience a substantiated estimate is very expensive and the number of failed initiatives does not decrease. What is causing this?
For a complex project a lot of experts are required
Many IT projects are realized within an existing application portfolio, so a lot of different expertise is required to shape a good solution. Quite often, the mistake is to rely on these subject matter experts to draw up the project estimate. Drawing up a total estimate for an IT project with subject matter experts requires a lead time of six weeks at the least. This approach also requires a good overview of the project to determine whether all parts have been estimated without overlap.
Metri’s research shows that this approach leads to an estimate that is 30% too optimistic on average in both cost and lead time. Subject matter expertise is no guarantee for a good estimate. Drawing up a good estimate is an expertise in itself. This requires sufficient knowledge to understand the architecture of a complex IT project, but above all knowledge of how this translates into an estimate. A good estimate is more than the sum of a number of separate activities. By using the right expertise, a higher quality estimate can be made in less lead time.
Every project is unique
The most often heard reason to rely on subject matter experts is that every project is unique and therefore requires subject matter knowledge to estimate it. The uniqueness defines the difference between a project and a recurring activity. But unique projects also adhere to laws of their nature. As a result, the estimate of a project is more than the sum of a number of separate activities. These activities must be planned, coordinated and controlled. Partial products must be integrated, tested and accepted. Estimation research has been conducted for over four decades and a number of patterns have emerged that also apply to the actual way projects are executed today.
External expertise is very expensive
Not every organization that executes IT projects has in-house estimation expertise. This is understandable, because today every organization uses IT facilities and almost every organization executes one IT project or more. If you want to do a proper estimate, you need to hire external expertise. That requires actual out-of-pocket spending. Sometimes as much as 1% of the total project budget. That seems like a lot of money to some organizations. But is that really true? What is the real cost to have subject matter experts spend six weeks on putting an estimate together? This cost is often not directly visible, because the subject matter experts are working on the project anyway. If you are able to make this cost transparent, an (external) estimation expert is not that expensive.
A bad estimate is more expensive
Moreover, an estimate put together by subject matter experts also entails a hidden risk, that a low-quality estimate entails additional costs for the project. It may sound strange, but a bad estimate simply costs money.
When an estimate is too optimistic, costs will increase exponentially. For example, because activities have to be executed in parallel that could better be executed sequentially. Or because people have to be added to the project to meet deadlines. This is relatively expensive and gives little return on investment.
But also a too pessimistic estimate costs money, although that usually remains hidden. When an estimate is too generous, the extra time or money is usually spent. For example, to build extra features that are not really necessary. The project could have been executed for less money or less time with the same value for your organization.
Interested in what a good estimate can bring to your organization? Please feel free to contact us. An exploratory meeting is completely free of charge.