.png)
We ran our Founder Studio - an accelerator for early-stage learning and work founders - for the first time from April to June. During the programme, we ran many sessions with founders, senior operators and corporates. This is the second of many articles we will publish built on the back of these sessions.
-----
If you ask an early-stage B2B founder how they close their first handful of enterprise customers and most will describe some version of the same motion: find a way to get in the door, walk through it, offer a pilot, let the product speak for itself, hope it converts. It feels generous. It feels low-friction. It feels like the obviously right way to earn trust before asking for money.
It is also, very often, how promising deals quietly die.
During the Studio programme, Andrea, our Customer Success mentor, delivered a great session, featuring her 'Pilots That Convert' logic. It makes an argument we think every founder running a pilot-led sales process should heed: a pilot is a structured way to de-risk a purchasing decision. If you treat it differently to this, you're offering free consulting with a demo.
A pilot is not a favour
Andrea's framework opens with a distinction that sounds almost too simple to matter, until you notice how often it's violated.
A pilot is a structured way to de-risk a purchase decision for a known use case. It's time-bound, outcome-linked, and decision-gated - the minimum proof needed to unlock a production go-ahead.
A pilot isn't free consulting, open-ended exploration, a feedback-gathering exercise, a mini-implementation with no end date, or a vague attempt to win buy-in.
Andrea's test for which one you're actually running is blunt: if the customer can't tell you what decision the pilot will unlock, it isn't a pilot yet. It's more a favour you're doing them and favours don't have a natural moment at which someone reaches an agreement for a contract.
This matters more for founders than for enterprise sales reps at incumbents, because founders have the least slack to absorb a pilot that goes nowhere. A three-month unpaid pilot that ends in "thanks, we'll think about it" can cost you more than a lost deal, particularly if you are looking to utilise the partnership to help you raise your Seed or Series A round.
The best pilot is the one you don't need
The most valuable part of the session may at first feel a little counter-intuitive: before you agree to run a pilot at all, Andrea's advice is to ask whether you need one.
Customer references, case studies with a quantified ROI model, a security packet shared proactively rather than requested, one architecture-review call with the prospect's IT team, a proper solutioning conversation with a sales engineer - all of these do the same de-risking work a pilot is meant to do, without the multi-month limbo. "The best pilot is the one you don't need" is a strange thing for a customer-success framework to say, but it's the right instinct. A pilot should be the tool you reach for when nothing lighter-weight will unlock the decision - not the default first move because it's what founders have seen other founders do.
If you do run one, run it like a contract, rather than a favour
Assuming a pilot really is warranted, the framework gives it a shape.
Four questions before kickoff:
1. What decision does this pilot unlock?
2. Who decides success?
3. Who signs the budget?
4. Who can block it?
That last one is the one founders skip most often - mapping procurement, security, IT, and legal before the pilot starts, because any of them can tank a deal after a technically successful pilot.
Success criteria have to be business outcomes - dollars saved, revenue gained, risk avoided, time-to-value reduced - with a baseline, a measurement source, and a decision date already on the calendar for scale, extend, or stop. "The tool connects to our data" is not success. Scope should shrink to the smallest viable proof: one wedge use case, two to three days of proof-of-value rather than weeks, hard guardrails against scope creep posing as customer responsiveness. And the path to production - security review, procurement, data access, legal - starts on day one or as early as possible, in parallel rather than after the pilot succeeds.
Andrea refers to this a Mutual Action Plan. Most pilots that drag aren't failing technically; they're waiting on a legal review that was most likely started too late.
Free pilots invite drift. Paid pilots invite decisions.
If there's a dichotomy in this framework worth putting front and centre of your considerations, it's that a pilot with no commercial or legal structure around it is an experiment nobody is obligated to conclude. A paid pilot, or a full MSA with a pilot period and an exit clause, forces budget approval before work begins, covers your delivery cost, and - most importantly - creates genuine seriousness on both sides.
Founders ought not to consider pilots as a lower-commitment way to start a relationship - it's supposed to be the opposite: a higher-commitment way to compress the time between "interested" and "decided," precisely because both sides have put something on the table.
Where this framework needs a caveat
Very early-stage, pre-product-market-fit founders often don't have the leverage to demand a paid pilot or a signed MSA from a first logo. You may not yet have the case studies, references, or track record that make "Step 0" alternatives credible substitutes for a pilot. In that phase, some willingness to absorb pilot risk in exchange for a design partner is a reasonable trade - the discipline should be clear in scoping and decision-gating even when it can't yet feed into the commercial terms.
The risk in the other direction is real too: too much process, too early, can feel like excessive bureaucracy to a prospect who just wants to try the thing. The structure here is meant to compress ambiguity, not add ceremony. If your go/no-go gate becomes a twelve-slide internal exercise before every deal, you've missed the point.
For founders building or scaling a GTM motion
If you're pre-PMF and running your first few pilots: borrow the discipline, but don't get tied down in paperwork. Name the decision the pilot unlocks and the person who'll make it, even if you can't yet insist on a signed contract before you start.
If you already have a founding CS or sales hire and are seeing pilots stall at the finish line, this is unlikely a product problem and more likely a commercial model problem. The pilot "worked," and it's dying in procurement or legal because these things take time and perhaps weren't started early enough.
If you're past a handful of pilots and still converting below plan: run Andrea's go/no-go gate as a filter before you agree to the next one. A pilot that fails the gate isn't a pilot yet - instead, it's a "pilot design" conversation you haven't had.
We think about GTM discipline as part of the operating muscle we look for in the companies we back, alongside product and team. If you're a founder wrestling with pilot-to-close conversion, or you're building tools or services that help B2B teams run this kind of structured de-risking better, we'd like to hear from you.
