Organizations are surprisingly good at saying yes and surprisingly bad at saying no. A promising idea gathers momentum, champions attach their credibility to it, sunk costs accumulate, and by the time the doubts surface, killing it feels like failure. So weak bets stagger forward, consuming resources that better opportunities needed. The problem isn't a shortage of ideas, it's a shortage of disciplined "no."
A research-led go/no-go process fixes this by making the kill decision evidence-based and early, before momentum makes it emotional. This guide shows how to build one.
The cheapest "no" is the early one. A bet killed at the research stage costs a study; the same bet killed after launch costs a year and a reputation.
Why companies can't say no
Saying no is psychologically hard for reasons that have nothing to do with the opportunity's merit. Sunk-cost bias makes already-spent resources feel like reasons to continue. Champion bias ties the idea to a person's credibility. Optimism reframes every red flag as a solvable challenge. And momentum makes stopping feel like quitting. Left unchecked, these forces keep bad bets alive long past the point where evidence said stop.
The antidote is to take the decision out of the realm of emotion and momentum and into the realm of pre-agreed evidence.
The research-led go/no-go process
A disciplined process puts evidence gates between an idea and full commitment:
- Frame the bet as a clear hypothesis with explicit success conditions.
- Identify the riskiest assumptions, the ones that, if wrong, kill the bet.
- Research those assumptions first, cheaply, before committing real resources.
- Test against pre-defined criteria at each gate.
- Decide go/no-go on the evidence, and genuinely mean the no.
The key move is sequencing: research the things most likely to kill the idea first, so a fatal flaw surfaces while it's still cheap to walk away.
Don't research to justify the decision you've already made. Research the one thing that could prove you wrong, first.
Test the riskiest assumption earliest. Most go/no-go processes fail because they validate the easy, comfortable questions first and only confront the deal-breaker after they're already committed.
Defining kill criteria upfront
The single most powerful discipline is defining kill criteria before you start, the specific conditions under which you will walk away. "If demand validation shows adoption below X, we stop." "If the serviceable market is under Y, we stop." Agreed in advance, these criteria are objective; agreed after the data arrives, they're negotiable, and momentum always wins the negotiation.
Pre-commit to your kill criteria. A threshold set before you're emotionally invested is honest; one set afterward will always bend to the decision you already want.
Killing bets without killing morale
A "no" shouldn't feel like a failure, it should feel like the system working. Reframe the goal: the purpose of a go/no-go process isn't to validate ideas, it's to allocate scarce resources to the best ones. A bet killed early frees people and capital for something better, and the discipline to kill weak ideas is what makes the strong ones well-funded. Celebrate good kills as much as good launches.
A worked example
An Indian SaaS company is excited about launching a vertical product for the domestic logistics sector. Instead of building first, the team defines a kill criterion upfront: "if fewer than 4 of 15 target logistics firms confirm they'd pay for this in discovery interviews, we stop." The riskiest assumption, real willingness to pay, not polite interest, gets tested first, cheaply. Nine interviews in, only one firm shows real budget intent; the rest like the idea but won't fund it. Because the threshold was set in advance, the team walks away in three weeks having spent the cost of a small study, not a year of engineering, and redirects that team to a bet that clears the bar.
Frequently asked questions
What is a go/no-go decision?
A structured decision point where an opportunity is either advanced or stopped based on evidence against pre-defined criteria, used to avoid over-committing to weak bets.
How do you decide to kill a project?
Define kill criteria upfront, research the riskiest assumptions first, and stop if the evidence fails those pre-agreed thresholds, rather than letting sunk cost and momentum decide.
Why is it so hard to say no to a project?
Sunk-cost bias, champion bias, optimism, and momentum all push toward continuing. A research-led process with pre-set criteria counters these forces with evidence.
When should go/no-go research happen?
Early, before major resources are committed. Researching the deal-breaking assumptions first means a fatal flaw surfaces while walking away is still cheap.
What makes a good kill criterion?
It's specific, measurable, and set before the research begins, a threshold like "if confirmed willingness to pay is below X, we stop." Vague criteria ("if it doesn't look promising") always bend to momentum; a pre-committed number stays honest when the data arrives.
Future outlook
As fast, affordable intelligence makes it easier to test assumptions early, the competitive advantage shifts to organizations with the discipline to act on a "no." In a world where research can de-risk decisions in days, there's less excuse than ever for letting a bad bet survive on momentum alone. The best portfolios aren't the ones with the most launches, they're the ones with the most disciplined kills.
The question for any opportunity on the table: have we defined what would make us walk away, and would we actually do it?
Key takeaways
- Companies struggle to say no due to sunk cost, champions, optimism, and momentum.
- A research-led process puts evidence gates before commitment.
- Test the riskiest, deal-breaking assumptions first.
- Pre-define kill criteria so the decision stays objective.
Want to kill bad bets before they cost you?
Zapulse runs rapid, research-led opportunity assessments that test your riskiest assumptions first, so you commit only to the bets worth making.
Talk to our analystsRelated reading: The Market-Entry Playbook · How to Size a Market: TAM, SAM & SOM
By Zapulse Research Team · Published Jun 15, 2026 · 7 min read · Strategy
