Discovery Should Change Decisions starts with a simple test: does this make the work more decidable, or does it only make the work easier to describe? Discovery Should Change Decisions: in discovery and customer research, teams often mistake fluency for progress. Discovery Should Change Decisions: they can explain the issue, name the stakeholders, and produce a tidy artifact while the actual research decision remains untouched.

Discovery Should Change Decisions matters because operating systems decay when decisions stay implied. Discovery Should Change Decisions: the company keeps moving, but each team carries a different version of the truth. Discovery Should Change Decisions: one group thinks the bet is strategic. Discovery Should Change Decisions: another treats it as optional. Discovery Should Change Decisions: a third waits for a signal that nobody has agreed to produce. Discovery Should Change Decisions: the surface looks aligned until execution exposes the disagreement.

Discovery Should Change Decisions is the part of the series that defines the central tension and names the work that the rest of the series will keep testing. Discovery Should Change Decisions: the post should therefore leave the reader with something more useful than a principle. Discovery Should Change Decisions: it should leave a sharper question, a better artifact, and a way to inspect whether the work changed.

Where the work breaks

Discovery Should Change Decisions breaks when the team keeps the conversation abstract. Discovery Should Change Decisions: abstract language lets everyone nod because nobody has to give anything up. Discovery Should Change Decisions: a real decision has a cost. Discovery Should Change Decisions: it changes priority, sequence, ownership, scope, customer contact, or follow-through. Discovery Should Change Decisions: if none of those things changes, the team may have had a good conversation, but it has not changed the operating system.

Discovery Should Change Decisions also breaks when teams use process as a substitute for judgment. Discovery Should Change Decisions: a meeting can collect updates without creating insight. Discovery Should Change Decisions: a memo can summarize context without recommending a choice. Discovery Should Change Decisions: a dashboard can show movement without showing whether the movement matters. Discovery Should Change Decisions: the repair is not more ceremony. Discovery Should Change Decisions: the repair is a clearer relationship between evidence and action.

Discovery Should Change Decisions has another failure mode: people protect optionality until the decision window closes. Discovery Should Change Decisions: optionality feels responsible because it avoids premature commitment. Discovery Should Change Decisions: past a certain point, though, optionality becomes a tax. Discovery Should Change Decisions: teams keep weak work alive, delay learning, spread attention thin, and make every downstream handoff harder.

What good looks like

Discovery Should Change Decisions is healthy when a team can say what changed after the conversation. Discovery Should Change Decisions: the change might be small: a narrower customer segment, a stopped feature, a clearer launch owner, a better research question, a different account plan, or a new review date. Discovery Should Change Decisions: small changes count when they remove ambiguity and create forward motion.

Discovery Should Change Decisions should make disagreement more useful. Discovery Should Change Decisions: good disagreement is not noise. Discovery Should Change Decisions: it is information about assumptions, risk, incentives, and evidence quality. Discovery Should Change Decisions: the operating move is to capture the disagreement in a form the team can test. Discovery Should Change Decisions: if the disagreement cannot be tested, it should at least be named as a judgment call instead of hidden as consensus.

Discovery Should Change Decisions should also protect the team from false completeness. Discovery Should Change Decisions: complete-looking artifacts can still avoid the most important question. Discovery Should Change Decisions: the test is whether a new person could read the artifact and understand the decision, the evidence, the tradeoff, the owner, and the next inspection point without reconstructing the whole history.

The useful artifact

Discovery Should Change Decisions needs an artifact that is small enough to survive normal work. Discovery Should Change Decisions: a useful artifact has five parts: the decision, the evidence, the tradeoff, the owner, and the review trigger. Discovery Should Change Decisions: anything beyond that should earn its place.

Discovery Should Change Decisions should name the decision in plain language. Discovery Should Change Decisions: if the decision is actually three decisions, split it. Discovery Should Change Decisions: if the decision has already been made, say that and use the artifact to clarify execution. Discovery Should Change Decisions: if the decision is still open, make the options visible enough that people can argue about the real choice.

Discovery Should Change Decisions should treat evidence with respect without worshiping it. Evidence has shape. Discovery Should Change Decisions: a customer quote, usage trend, sales objection, churn pattern, or support signal can matter a lot, but each proves a different thing. Discovery Should Change Decisions: the artifact should say what the evidence supports, what it does not support, and what would be strong enough to change the next move.

Discovery Should Change Decisions should make the tradeoff impossible to miss. Discovery Should Change Decisions: tradeoff language is the difference between a strategy document and an aspiration document. Discovery Should Change Decisions: the team should know what receives less capacity, what waits, what gets cut, what risk is accepted, and which stakeholder will feel the cost.

How to inspect it

Discovery Should Change Decisions can be inspected with four questions. What are we choosing? What are we refusing? Discovery Should Change Decisions: what evidence would change our mind? Discovery Should Change Decisions: what happens before the next review? Discovery Should Change Decisions: if a team cannot answer those questions, the work is not yet ready for more process. Discovery Should Change Decisions: it needs clearer judgment.

Discovery Should Change Decisions should show up in the calendar. Discovery Should Change Decisions: if the decision matters, it deserves a checkpoint. Discovery Should Change Decisions: that checkpoint does not need to be heavy. Discovery Should Change Decisions: it needs a defined signal, a real owner, and permission to change course. Discovery Should Change Decisions: without that, the team will keep carrying the decision as background anxiety.

Discovery Should Change Decisions should reduce the need for executive translation. Discovery Should Change Decisions: a senior leader should be able to inspect the work without redoing the thinking. Discovery Should Change Decisions: if the leader has to infer the customer, rebuild the evidence, guess the tradeoff, or identify the owner, the artifact is not doing enough operating work.

Field test

Discovery Should Change Decisions can be tested on one live piece of work this week. Discovery Should Change Decisions: pick something already consuming attention. Discovery Should Change Decisions: rewrite it as a decision, not a status update. Discovery Should Change Decisions: name the owner, the evidence, the tradeoff, and the review trigger. Discovery Should Change Decisions: then ask what changed because the artifact exists.

Discovery Should Change Decisions passes the test when the next action becomes more specific. Discovery Should Change Decisions: the next action may be a customer call, a killed initiative, a narrower scope, a pricing review, a product bet, a launch decision, or a management conversation. Discovery Should Change Decisions: the important part is that the work leaves the realm of explanation and re-enters contact with reality.

Evidence note: This is an operator-judgment essay grounded in Antoine's local source pack for Discovery and Customer Research Without Theater and adjacent series context, including https://www.antoinebuteau.com/building-ai-products-is-not-prompt-decoration/.

This is part 1 of 10 in Discovery and Customer Research Without Theater.