Innovation NavigatorStrategic Game
Test the belief your next build rests on before you build it: pick the riskiest assumption, design the cheapest credible test, write the threshold and what KILL means before it runs, and decide within 48 hours of the result.
Teams about to build something expensive on an untested belief: pre-launch tests, post-prototype gates, off-roadmap experiments, new products or markets, and any investment decision that keeps being argued from opinion.
From build it and see to set the kill line first.
We tend to read evidence in the direction we want it to go, so the threshold is written before anyone sees the data. Deciding in advance what will trigger a change tends to guard against staying the course because of what’s already invested. In a randomized field trial, founders taught to state hypotheses and test them before committing were more likely to drop or change weak ideas.
List the beliefs the plan rests on, mark each for importance and evidence, and test the one that matters most with the least evidence. In a randomized field trial, founders who tested their hypotheses first were more likely to drop or change weak ideas.
The cheapest method that passes three checks (behavior not opinion, a comparison, a big enough sample), run honestly with the people in it. A test that fails a check is a signal for a bigger test, not a decision to scale.
KEEP IF, KILL IF and what KILL means, written and signed before the test runs; the decider calls it within 48 hours of the result. We tend to read evidence the way we want it to go, and deciding the trigger in advance tends to guard against staying the course because of what’s already invested.
Also runs as a half day · P4.7 HD
Half day (3 h 30 min), “Ideate, build, test”: P4.10 Lean Ideation in the morning, then this lab’s test card written before anything is built, a prototype tried on 3–5 guests from outside the team, and a review after the real test.
Take one sharp question to a tested prototype in a single session: generate alone before choosing on criteria, write the test card before building anything, build only what the test needs, and try the prototype on people outside the team while the test is still cheap to fix.
Product, innovation and strategy teams with one sharp problem and a build decision coming: innovation sprints, working off-sites, and teams that tend to settle on the first plausible idea or build before they test.
People working alone and then pooling tend to produce more ideas than the same people brainstorming aloud together, and when choosing, people tend to pass over their most original ideas, so the pick gets checked. We tend to read evidence in the direction we want it to go, so the threshold is written before anyone sees the data, and deciding in advance what will trigger a change tends to guard against staying the course because of what’s already invested.
Everyone generates in silence before pooling, constraint rounds push past the obvious, and the pick is made against criteria agreed before the vote, with the safest-idea check. People working alone and then pooling tend to produce more ideas, and people tend to pass over their most original ones when choosing.
The riskiest assumption, a credible test, and KEEP IF, KILL IF and what KILL means, written before anything is built; the prototype is only what that test needs. We tend to read evidence the way we want it to go, so the line comes first.
Guests from outside the team use the prototype while observers note what they do; the call is RUN, REVISE or SWAP, and the idea waits for the real test. Deciding in advance what will trigger a change tends to guard against staying the course because of what’s already invested.
The thinking behind this lab, published in full: