Every transformation project has its ROI slide. Numbers in a table, assumptions in a footnote, results promised in 18 months. Boards approve it, the CFO signs off on it, project managers keep watch over it.
And yet most transformations fail. Not because the numbers in the table were wrong, but because no one measured what has to change before the numbers can move at all.
People's behavior.
What ROI doesn't see
ROI is an output metric. It measures the result after something has already happened. It's a rearview mirror — useful, but late.
The problem with transformation projects isn't that companies can't calculate a return. The problem is that there's a gap between the decision and the result that no financial model captures: the time it takes for new behavior to actually become part of everyday work.
One of our clients, a mid-sized financial institution, invested in a new sales system and a six-month training program. The ROI model promised a return within fourteen months. A year later, sales results had barely changed. The system was deployed, people completed the training, reports showed activity. But the way salespeople actually ran meetings with customers stayed exactly the same as before the project.
ROI never caught that. Because ROI doesn't ask about behavior.
Time-to-behaviour-change: what it is, and why it matters
Time-to-behaviour-change is a metric that measures how long it takes for a new, desired behavior to become a standard part of the work routine — without conscious effort, without reminders, without external pressure.
It isn't an activity metric. It isn't enough for people to open the tool. It isn't enough for them to complete training. Time-to-behaviour-change measures the moment a new way of working becomes automatic — the moment “I should try this” turns into “I just do this naturally.”
Why does this metric matter more than ROI?
Because it's causal, not correlational. Shortening time-to-behaviour-change directly drives better results. Improved ROI is a consequence, not a cause. A company that knows how to manage this time knows how to manage its results. A company that only tracks ROI is only tracking the consequences of decisions it already made — or failed to make — a long time ago.
What time-to-behaviour-change looks like in practice
Picture two scenarios for rolling out an AI tool to help salespeople prepare for customer meetings, both of which we've seen with clients:
Scenario A: The company buys a license, IT deploys the tool, HR prepares an e-learning module, managers get a presentation. After three months, the adoption rate is 34%. After six months, 41%. The project gets labeled a success.
Scenario B: The company defines a precise context of use — the tool opens as the very first step of every meeting prep, not “whenever it's convenient.” The first ten salespeople get individual feedback within the first two weeks. Results from the first week are visible immediately. After three weeks, the adoption rate is 78%. After six weeks, the behavior is stable, with no reminders needed.
Both scenarios share the same ROI assumption. But the time to behavior change differs by more than three months. And three months, in a sales environment, means lost opportunities and a slower sales process, because people keep operating the old way inside the new system.
Three factors that most affect time-to-behaviour-change
Our experience across dozens of transformation projects shows that behavior change isn't chance. It's the result of three variables that can be actively managed.
1. The precision of the trigger context. The more specifically you define when and where the new behavior should occur, the faster the brain forms an automatic association. “Use AI when preparing for meetings” produces change on the scale of months. “Open the tool whenever a calendar invite arrives” produces change on the scale of weeks. The difference isn't people's motivation — it's the architecture of the trigger.
2. The speed of the first feedback. The brain reinforces behavior that produces a visible result. The longer the loop between the new behavior and the feeling that it makes sense, the higher the risk that people drop off before the habit forms. Behavior-change projects that take less than four weeks generally share one thing: the first concrete result is visible within 72 hours of launch.
3. Social normalization. An individual can adopt a new behavior. But if the team around them still operates the old way, the social norm will pull them back regardless of motivation or conviction. The time needed for change shortens significantly the moment the new behavior is adopted first by the people others already watch and follow — not necessarily managers, but the informal authorities on the team. Change ambassadors.
What this means for company leadership
If you're a CEO or CHRO approving transformation projects, ask yourself three questions the next time you make a decision:
1. How long until we see the first proof of behavior change — not activity, but real change? If the answer is more than six weeks, the project probably doesn't have a sufficiently radical behavioral layer built into it.
2. Who specifically is accountable for the behavior change — not for the adoption rate, not for the number of training sessions completed, but for the new behavior actually becoming automatic? If the answer is “the project manager” or “no one specifically,” you have a blind spot. You need to find an adoption owner at the business level.
3. How will we know the behavior has changed, and exactly how will we measure it? If the answer is “we'll see it in the results a year from now,” you're measuring the consequence, not the cause.
Why this is becoming an existential question right now
At a time when companies are rolling out AI tools broadly and quickly, time-to-behaviour-change is becoming a critical variable of competitiveness.
Technological capabilities are available to everyone roughly equally. LLM models still cost next to nothing. The difference between a company that genuinely profits from AI and one that pays for it without results isn't which tool it bought. It's how quickly it can anchor new behavior into everyday work.
Companies that know how to consciously shorten time-to-behaviour-change are building a real competitive advantage that doesn't live in systems, but in people. And that's considerably harder to copy than software.
ROI will vary by percentage points. Time to behavior change can vary by months. And at a time when markets are changing faster than ever, time matters more than percentage points.
Kogi measures and shortens time-to-behaviour-change at companies. If you're curious how that would work for you, get in touch.