Return on Investment (ROI) is the most common metric used to stack-rank projects for funding and the “go” decision.
Unfortunately, for most organizations, the ROI promised is seldom what’s delivered.
Let’s begin with a fact of organizational life. Ideas almost never get acted upon immediately.
The enterprise has a planning cycle: little to nothing gets injected in the middle of a plan year. Nothing much happens, in other words, until the next plan cycle at the earliest.
Plan cycles, in turn, don’t start right away, either. Most organizations begin planning about now — July — for a year that won’t begin until six months from now, in January. So the idea is stale by at least half a year simply due to the process.
Queues and Backlogs
Just because the idea is good doesn’t mean it will be funded and selected in the first plan cycle that passes.
It can be queued for a later-in-the-year start. It can remain on the backlog, either because there are other projects with “better ROI” ahead of it, or because other projects have been deferred too long, and must go now (principle of equity), or because compliance or vendor-induced change takes precedence.
Few enterprises revisit business cases and assumptions while items are queued or on the backlog. External changes, therefore, are seldom taken into account.
Those that do undertake a second “go” decision — typically in government, where the Treasury Board (Federal) or Management Board (Provincial) reviews all expenditures again to release the funds — seldom question what the effects of backlogs and queueing has done, even when they do ask for updated cost estimates.
Third, every project manager knows that no project plan survives long in contact with reality. At first, what’s likely to slip is the delivery date. Later, scope will be dropped in order to complete. In both cases, cost continues to ratchet upward.
A surprising number of projects lose their ROI — falling below the “clip level” below which no project would be approved, or in some cases actually “going negative” — through these delays. The “payback” point (where the benefits that accrue recover the costs of achieving them) certainly moves out — and it’s not uncommon for them to move two, three or more years beyond when they were expected.
Once the project is delivered, it must actually have an effect in the workplace or marketplace. Almost no organizations expend effort here, to find out whether or not services that were planned are actually used. Insufficient measurement is done, to determine if the solution is making a difference.
The Governance Board’s Responsibility
The role of the IT Governance Board is in large measure to generate returns on investment. How should it go about that?
First, keep the ROI up to date. Queue ideas, not approved projects. It’s far cheaper and easier to keep updating plans and costs than it is to release them, locking them into place, then trying to revise them after the fact.
Second, be unabashed about freezing or terminating projects that go well off the rails. If the ROI is lost, it’s not going to be earned back, so the question needs to be “is there something else we could do now that would give us some return?”
Third, insist on projects including a deployment phase that optimizes use and/or measures reactions. These are easy fixes to ensure the ROI comes closer to happening once the product is usable.
Fourth, when prioritizing, include some of the infrastructure or update projects each year, even though their ROI is lower. These almost always achieve their objectives, giving a floor return to work with for the enterprise and keeping the IT base closer to the lowest cost to own and operate over time. After the initial transition to good governance, there should be no more surprise projects where a superannuated package must now be updated!
Real ROI is the only kind that flows to the enterprise’s bottom line. Govern accordingly.