Most organisations treat ROI like a passport stamp; you need it to get through approval, and up until that point, it genuinely feels like nothing could be more important, but once the project gets that all-important green light, no one ever really checks it again.
The spreadsheet is filed away and gathers dust, assumptions quietly age out, and delivery takes over. As most orgs love the good news and conveniently underplay that awkward bad news, if the project succeeds, then lo and behold, the business case gets the credit, whereas if it struggles? Let’s just look the other way and politely forget that inconvenient ROI forecast, whose total inaccuracy could make people look bad.
But the anecdotal evidence we’ve had from our customers is that something interesting happens when ROI doesn’t disappear after approval.
When teams know their assumptions will be revisited, compared to reality, and discussed openly, they build business cases very differently. And it’s not because they suddenly become any better at forecasting, but because their incentives change.
So join me in an adventure about behavior, not maths.
The hidden contract inside every business case
Every business case contains an unspoken agreement; approval teams agree to fund the initiative based on a set of assumptions, with delivery teams agreeing, implicitly, that those assumptions are “reasonable enough” to justify starting. But what rarely happens is a second conversation, where there should be questions such as:
Were the benefits actually realised?
Did the costs land where expected?
Were delays or adoption issues foreseeable?
Which assumptions held up, and which collapsed on contact with reality?
Without that second conversation, business cases become one-way artefacts, as they exist purely to unlock funding, not to inform learning. And people aren’t stupid, so when that’s the case, people quite understandably optimise for approval, not accuracy.
Why most ROI forecasts drift toward optimism
Now, this isn’t actually because people are inherently dishonest; it tends to be because forecasts live in this strange accountability vacuum. If ROI is never revisited, then optimistic assumptions come with little, if any risk, conservative estimates feel like self-sabotage, and complexity becomes something to smooth over, not confront.
Effectively, it means that no one is rewarded for being “mostly right but cautious”, instead, they are rewarded for getting a project approved. So it’s quite rational (and in no way malicious) that teams lean into aggressive benefit ramp-ups, best-case adoption curves, and soft dependencies that never quite make it into the numbers.
It’s not intended, and certainly not desired, but the system quietly teaches people what really matters, and that’s definitely neither underplaying things nor accuracy.
What changes when ROI is tracked after approval
Now, picture a scene; we’re in a different environment, and rest easy as it’s not a punitive one, just one where ROI is still visible six, twelve, eighteen months later, and where assumptions aren’t used to judge people, but to understand outcomes.
It didn’t take much, but suddenly we’ve reached a magical world where the business case stops being a sales pitch and starts behaving more like a hypothesis. And you know what’s even more fun? It doesn't actually take too much magic to get there and it isn’t just a hypothesis, it’s the feedback we’ve had from customers, who have told us that it also creates some very welcome behavioural shifts…
🕵️♂️ Assumptions become explicit, not implied
If you know assumptions will be revisited, you surface them. This means that adoption timelines get written down instead of hand-waved, dependencies get named, and risks move from footnotes into the core narrative. And this isn’t happening because someone demands it, but because vague assumptions tend to sit fairly uncomfortably when you expect to see them again.
🎢 Confidence replaces false precision
Teams stop chasing spurious accuracy, so instead of the incredibly precise-sounding (yet likely entirely untrustworthy) “Benefits grow by 23.74 percent in year two”, you start to hear people talking about ranges, scenarios, and breakpoints. The conversation moves from:
“Is this number correct?”
to:
“What would have to be true for this to work?”
and that’s a far more useful question for all concerned.
🤹 Trade-offs get discussed earlier
When ROI lives beyond approval, those handy little shortcuts that people know will cause problems, but they take anyway because it’s not like anyone will ever know, suddenly become visible. You know, the classics: deferred change management, underestimated effort, “Phase two” benefits with no owner.
These things are a lot harder to bury when you know the story doesn’t end at go-live, which means that you hear teams more willing to say things like:
“This delivers less upside, but it’s more resilient”
“This risk materially changes the outcome”
“We can approve this, but only if we invest properly upfront”
and again, those are far healthier conversations, even when they lead to tougher decisions.
The real value isn’t control, it’s learning
There’s a fear that tracking ROI post-approval turns into surveillance, and in short-sighted orgs, that’s unfortunately an entirely justified fear, as it wrongly becomes a stick rather than a lens. But those orgs that are smart enough to actually practice growth and learning, rather than just endlessly talking about it, it isn’t used for blame, it’s used for memory.
In all honesty, the vast majority of organisations are terrible at remembering, with the main offenders being which assumptions tend to fail, where benefits consistently lag (and they are therefore completely oblivious about why), how long adoption really takes, and which types of projects routinely overpromise.
In corporate language, you could say that without feedback loops, every business case is built in isolation. Or as I would say, “Fool me once….”.
When you bother reviewing how past projects played out though, you’ll start to find that patterns emerge. And when you choose to learn from and act on those patterns, you find that teams get better, not because they’re punished for being wrong, but because they’re finally allowed to learn from being wrong.
Why this improves decisions, not just forecasts
The irony is that the feedback we’ve received says that post-approval ROI tracking often leads to fewer inflated numbers upfront. And that’s not because teams are forced to lowball, but because realism becomes socially acceptable.
When decision-makers start to value insight over optimism, they tend to find that:
Killing a weak project feels responsible, not risky
Scaling back scope is seen as discipline, not failure
Revising forecasts is treated as stewardship, not backtracking
ROI stops being a performance claim, instead becoming a decision tool again (which is what it was always supposed to be…)
A quieter cultural shift
The most interesting impact of tracking ROI after approval we’ve heard, though, isn’t in dashboards or reports, it’s in how people talk. Suddenly, you tend to find that there are fewer heroic promises, more honest uncertainty, better framing of risk, and a clearer ownership of outcomes.
Business cases become less about “selling the future” and more about navigating it. And over time, that changes how organisations decide what’s worth doing in the first place.
The uncomfortable truth
If you don’t plan to look at ROI again, you’re implicitly telling people it doesn’t really matter, as approval is the finish line.
But if you do plan to look again (even lightly, however imperfectly), you send a different signal, as you tell teams that assumptions matter, learning matters, and decisions don’t end when funding is approved.
And you know what? That signal alone is enough to change behaviour (and it didn’t even require a governance overhaul…)
