A large prospect asks for six features, a security review and a free pilot. The logo would look excellent in the deck. The founder spends a month responding before learning that no budget owner has joined, no deadline exists and the current process is merely annoying.
Founding-sales practice keeps experimentation separate from scaling. Question-led selling uncovers the consequence and buying path, while evidence-safe interviewing keeps a founder from manufacturing polite agreement. Together they create a qualification spine around fit, pain, process and reciprocal commitment without pretending the sales recipe is finished.
Qualify evidence, not enthusiasm
A deal advances on customer commitment, not enthusiasm.
01 · Fit
Match the case
Is this the company and user you can serve now?
02 · Pain
Find consequence
What breaks, costs or remains blocked today?
03 · Path
Map the decision
Who owns budget, risk and final approval?
04 · Commitment
Ask for a step
Will the customer invest time, data or money?
Use four tests before adding a deal to the forecast
Fit asks whether the organisation and user resemble the group the product can help now. Pain asks for the current consequence and why it matters this quarter. Path identifies the people, evidence and approvals required to buy. Commitment asks the customer to take a proportionate next step. Excitement is useful; reciprocal action is stronger evidence.
- Pursue: fit, consequence, path and commitment are credible.
- Learn: the case is outside today's ICP but tests a named market or product belief.
- Stop: no material pain, owner or next step exists, and there is no deliberate learning reason to continue.
Make a pilot a buying test
A useful pilot names the problem, scope, customer owner, evidence supplied, success measure, decision date and commercial next step. Payment is not always possible, but some reciprocal commitment should be. A free sandbox with no owner or decision merely postpones qualification while creating support work.
Use AI to maintain the evidence trail
An agent can prepare a call around missing qualification facts, connect claims to the exact interaction and draft the agreed pilot. It should not manufacture a score from job titles or treat a polite reply as intent. Every stage change should carry the customer action that justified it.
Train the agent to earn the next sales step
A useful sales agent should be able to explain why a deal deserves the next unit of founder attention. Give it the customer record and qualification policy, then make every recommendation point to an observed customer action or a clearly labelled unknown.
- Evidence pack: approved ICP and exclusions, current product truth, call notes and transcripts, email and meeting history, opportunity stage, security or procurement requests, prior wins and losses, and the customer actions behind any existing forecast claim.
- Working brief: build a four-part evidence table for fit, consequence, buying path and commitment. Identify what is known, what is merely said and the one missing fact that should determine the next conversation.
- Permitted action: prepare the call brief, research missing stakeholders, draft the follow-up and turn an agreed problem into a scoped pilot with an owner, success measure, decision date and commercial next step.
- Review gate: a person approves sends, price, concessions, commitments and material stage changes. The agent must not convert seniority into buying power or move a deal because the conversation sounded positive.
- Learning record: record the exact customer action that advanced, paused or ended the deal, plus the founder's correction to the recommendation. Use won and lost outcomes to revise the qualification policy, not the raw CRM record itself.
This trains the agent away from persuasive summaries and towards the company's real standard of evidence. Over time it learns which customer actions predict progress and which polite signals merely consume attention.
A useful next move
Review the ten open deals consuming the most founder attention. Label each pursue, learn or stop. For every “pursue”, write the consequence, buyer path and reciprocal next step in one sentence. If you cannot, the next call is for qualification, not demonstration.
Continue with why the founder owns the first enterprise sales, or read when to hire the first salesperson.