Pivot, Persevere or Stop: How to Make the 3P Decision Without Fooling Yourself
Every innovation experiment ends with the same uncomfortable moment: you have data in front of you, and you have to decide what to do with it. Keep going. Change direction. Or accept that this idea isn’t working.
In 20 years of coaching innovation teams in FMCG, this is the moment I’ve seen go wrong more often than any other. Not because people lack intelligence — but because they’re human. They’re attached to their ideas, worried about what stopping will look like, and unsure how to read imperfect data.
This article is about how I think through that decision — and how to run the meeting that makes it less painful.
What the 3P Decision actually is
In the Lean Startup framework, after every Build-Measure-Learn cycle, a team must choose one of three paths:
- Persevere — the data confirms your hypothesis. You continue, possibly with refinements.
- Pivot — the data shows your strategy is wrong, but your underlying insight might still be right. You change a fundamental element of the approach.
- Stop (Perish) — the data shows no viable path forward. You close the project intentionally, before it drains more resources.
Simple in theory. Hard in practice — because none of these decisions feel clean when you’re inside them.
The most common mistake: staying too long
In corporate FMCG environments, the default is almost always to Persevere — even when the data says otherwise. I’ve sat in review meetings where a product concept had failed three consumer tests in a row, and the team was still arguing for one more round.
There’s a name for this: the Sunk Cost Fallacy. The logic goes — we’ve already invested six months and a significant budget into this. Stopping now would mean that was wasted. So we keep going.
But the cost is already spent. The only question left is whether to spend more. And that question has to be answered based on what the data is telling you — not what you hoped it would say.
When to Pivot vs. when to Stop
The decision between Pivot and Stop is the hardest one, and there’s no formula that removes the judgment from it. But there is a useful distinction:
- Pivot when: the quantitative data is weak, but qualitative feedback reveals that people genuinely care about the underlying problem — they’re just not responding to your current solution. The insight is valid. The execution is wrong.
- Stop when: both the quantitative and qualitative data are consistently pointing in the same direction — low interest, no emotional pull, and no clear signal that a different approach would change anything.
The other indicator I use: after how many loops does the team feel energized versus depleted? A genuine Pivot brings new energy. Teams that are just delaying a Stop tend to go through the motions.
How to run the 3P decision meeting in 3 steps
The meeting itself matters as much as the data. Here is how I structure it.
Step 1 — Restate the original hypothesis.
Before showing any data, go back to what you were trying to learn. “We hypothesised that working parents in France would pay a premium for a single-serve yogurt that required no spoon.” Anchoring on the original question prevents the meeting from drifting into justifications.
Step 2 — Show the data you actually have, not the data you wish you had.
Present the metrics that were defined as success criteria at the start — not the ones that happen to look good. If the success criterion was 60% purchase intent and you got 38%, show 38%. The temptation to lead with encouraging secondary data is where decisions go wrong.
Step 3 — Force a decision, and define what happens next.
The worst outcome of a 3P meeting is leaving with no decision. If the team chooses Pivot, the new hypothesis must be written before anyone leaves the room. If the team chooses Stop, a post-mortem must be scheduled within two weeks to extract the learnings formally.
The Zombie Project problem
The most expensive innovation failure isn’t a clean Stop. It’s a Zombie Project — one that isn’t succeeding but isn’t quite dead enough to be killed. It stays on the roadmap, consumes a small slice of budget every quarter, occupies good people’s time, and blocks new ideas from getting attention.
Zombie Projects survive for three main reasons: there are no pre-defined success criteria, leadership’s ego is attached to the idea, or the team is afraid that stopping will be seen as failure.
The antidote is simple but uncomfortable: define what “good enough to continue” looks like before the experiment starts. Not after. And treat a clean Stop — when the data warrants it — as evidence of good judgment, not a poor outcome.
“The best innovation decisions I’ve made were not the ones where I pushed the hardest. They were the ones where I read the data honestly and had the conversation nobody wanted to have.“
Amandine Devergies
If this resonates
I’ve applied this framework across dairy, frozen food, coffee, and cereals — each time with different data and different teams, but the same underlying discipline. If you’d like to see how it plays out in practice, the projects section walks through several real examples. Or connect with me directly on LinkedIn.
