GO / NO-GO decision framework for business ideas

A GO / NO-GO decision framework is a written contract that connects a business investment to the evidence, economics, risks, and deadlines that justify continuing or stopping. The result is not a confidence score. It is a decision, the conditions attached to it, and the next action that can change it.

The four useful outcomes

• —GO: the critical assumptions have enough evidence, the economics clear the required threshold, and no unresolved risk can invalidate the investment.
• —Conditional GO: proceed only with a bounded pilot, spending cap, deadline, and pre-agreed success and stop conditions.
• —Redesign: the problem may be real, but the customer, offer, channel, price, or operating model must change before more capital is committed.
• —NO-GO: a critical assumption failed, the economics cannot work within realistic constraints, or the remaining uncertainty is too expensive to resolve.

Write the decision contract first

Before research begins, record the decision owner, amount at risk, deadline, target customer, geography, expected price, minimum acceptable economics, and the three assumptions most capable of killing the idea. This prevents a team from moving the goalposts after weak evidence appears.

Minimum decision contract

• —Decision: the exact commitment being considered, not a general wish to explore.
• —Evidence threshold: what must be observed and from which source.
• —Economic threshold: price, gross margin, acquisition cost, payback, or another metric relevant to the model.
• —Kill criteria: the fact, result, or missed deadline that ends or redesigns the current version.
• —Budget and deadline: the maximum money and time allowed for the next proof step.
• —Owner: one person accountable for declaring the outcome on the agreed date.

Evaluate evidence in the right order

1. 01Confirm the customer and problem with direct first-party evidence.
2. 02Document current alternatives, switching costs, and the reason to change now.
3. 03Test willingness to commit through a paid pilot, deposit, signed letter of intent, procurement step, or equivalent behavior.
4. 04Bound the reachable market using the chosen geography, segment, channel, and delivery capacity.
5. 05Model unit economics with ranges and identify which variable can break the model.
6. 06Check regulatory, technical, operational, and intellectual-property constraints.
7. 07Compare every result with the decision contract and declare one of the four outcomes.

Set kill criteria before the test

A kill criterion must be observable, time-bounded, and linked to a critical assumption. “Customers seem interested” is not a criterion. “Fewer than three of 20 qualified buyers begin a paid pilot at $1,000 within 45 days” is testable. Missing the threshold does not prove the entire market is impossible; it means the current customer, offer, price, channel, or timing did not earn the next investment.

Worked example: B2B invoice automation

The numbers below are illustrative assumptions, not Trust Engine customer data. A team considers spending $60,000 to build invoice-exception software for U.S. distributors. Its critical assumptions are that finance teams experience a costly weekly exception problem, can approve a $1,500 monthly product, and can connect the required data without a long enterprise integration.

• —Conditional GO threshold: five qualified design partners, three paid 60-day pilots, and at least two integrations completed within ten business days.
• —Economic threshold: an evidence-supported path to 70% gross margin and customer-acquisition payback within 12 months.
• —Kill criterion: fewer than three paid pilots after 30 qualified conversations, or a mandatory integration that pushes delivery cost above the margin threshold.
• —Decision: fund only the bounded pilot first; do not release the full build budget until the pilot thresholds are met.

Common decision errors

• —Using a global market-size slide as proof of reachable demand.
• —Counting compliments, survey intent, or free trials as equivalent to payment or procurement behavior.
• —Hiding weak evidence inside a blended score instead of exposing the failed assumption.
• —Changing success thresholds after results arrive.
• —Choosing GO because research money has already been spent.
• —Declaring NO-GO without stating which changed assumption would justify a new test.

Move from verdict to action

Start with the guide to validating a business idea before spending money, use these official U.S. market research sources, review the Trust Engine methodology, or request a proposal for a specific investment decision.