Skip to main content

Paid pilot scope and success criteria: agree these 4 first

Paid pilot scope and success criteria: agree milestones, named owners, acceptance evidence and an end-date decision before your startup's pilot begins.

Michael Houck6 min read

Before a paid pilot starts, a startup founder and the customer's champion should agree in writing on 4 things: the milestones and what each side delivers for them, a named owner on both sides, the evidence that counts as success, and the end date with the decision that follows it. Put those terms in the purchase order or memorandum of understanding, not a loose email thread. Vague milestones and no named owner are 2 of the reasons pilots stall, according to Pepe Agell's March 2026 guide for Pear VC.

ResiQuant moved from free POCs to paid ones with fixed terms

Omar Issa, co-founder and CEO of ResiQuant (PearX S24), described the switch in Pepe Agell's March 19, 2026 guide to running pilots for Pear VC. "Early on at ResiQuant, free POCs were the right call," Issa said there, because they helped validate product-market fit with design partners during PearX. As the company scaled after its seed round and sold to insurance carriers and MGAs, he said, "Pepe’s framework helped us transition to paid POCs with clear milestones and fixed timelines."

Issa's account of the result, in the same Pear guide: customers "now consistently complete POCs fully integrated and production-ready," and "customers assign dedicated resources, we commit senior engineering time, and pilots stay on track." That is one CEO's description of his own sales motion, not a controlled study. It is still a specific picture of what the structure is supposed to buy you: a customer who staffs the pilot and a pilot that finishes.

The 4 agreements to write down before day one

Milestones, with deliverables on both sides. Agell's March 2026 Pear guide recommends naming the top 2 to 3 milestones, the deliverables from your team and from theirs, and a realistic timeline for each step. Write the customer's deliverables as plainly as your own: data access by a date, a sandbox, a user group, a weekly slot on someone's calendar.

A named owner in each company. The same Pear guide asks for a clear owner of the pilot inside the customer organization plus a counterpart on your side. Use names, not departments. Agell lists a missing internal owner among the reasons pilots stall, alongside vague milestones and work that feels optional because the pilot is free.

The evidence that counts as success. Agell's March 2026 guide says a pilot should move forward on 3 fronts: the integration and workflows are set up, the customer sees early results that clarify your value, and, if pricing depends on consumption, both sides learn what normal usage looks like. Turn each front into a sentence someone can check: the champion can show the before and after in their own report, which is far more useful than a general sense that the team likes it.

An end date and the decision after it. Agell's March 2026 guide says to aim for 1 month and keep a pilot under 3 months no matter what. Book the decision meeting before the pilot starts, and write down what happens at it: convert, end, or extend once on new terms.

The price still matters, and the same Pear guide is specific about it. In that March 2026 guide, Agell suggests a pilot price that covers your costs, sits roughly in line with a prorated slice of the annual contract, and carries a discount of 20% to 50% to lower friction. His guide also suggests discussing the annual contract value before the pilot begins and crediting the pilot fee toward that contract if the customer signs within a set window after the pilot ends. If you're still working out what one finished result is worth, my free Sell the work skill walks you through defining one unit of work, what it costs you to deliver and a floor price, which is the same thinking a pilot price needs.

What my first paying users taught me about acceptance criteria

Megaphone's first month is the closest thing I have to a pilot record: I priced it at $1,000 a month and added 3 users in that month, as I wrote in Why You Should Charge Users Early in January 2025. Those 3 users each worked closely with me through that month, with dedicated Slack channels and detailed daily feedback, and we kept all 3, according to that same January 2025 essay.

That wasn't a formal pilot, and I'm not claiming it was. The part that carries over is what the 3 users turned into: each was different, and those 3 became our 3 customer personas, per the same January 2025 essay. If your first 2 or 3 pilot customers are different kinds of buyers, give each its own success criteria. Averaging them into one checklist hides the fact that you're learning about 3 businesses at once.

Paying changes the conversation for the same reason. In that essay I argued that money on the line makes customers more candid about what you need to deliver. To hear it while it can still change the pilot, book the feedback review with your champion before day one.

Where this advice doesn't apply

Free pilots can still be right. Issa himself said free POCs were the right call for ResiQuant early on, while it was validating product-market fit with design partners, according to the Pear guide. If you don't yet know what problem you solve or for whom, you're running discovery, and a paid pilot with acceptance criteria is premature.

The structure here is built for B2B products where a customer has to integrate something and assign people to it. A self-serve consumer app or a cheap monthly tool usually doesn't need a pilot at all. And none of this is legal advice: if the pilot touches customer data, security reviews or procurement terms, get those reviewed separately.

AI tool pilots in particular often stop short of a rollout, which I wrote about in AI Pilots Are Failing in September 2025, using an MIT report. Written criteria won't fix a weak product, but they do tell you sooner whether you have one.

Dreamit Ventures managing partner Steve Barsh covers similar ground in a short talk, with chapters on clear success criteria, measurable KPIs, a defined time frame, and pilot pricing and conversion:

A pilot acceptance sheet to send your champion

Copy the sheet below, fill it in with your champion, and attach it to the purchase order:

  1. Pilot name, start date, end date, decision meeting date and attendees.
  2. Milestones, each with an owner on our side, an owner on yours and a due date.
  3. Customer deliverables: data access, environments, users and meeting time, each with a date.
  4. Success evidence, each item written as something one of us can check and naming who signs off.
  5. Price, what it covers, and whether it is credited toward an annual contract signed within an agreed window.
  6. Annual contract terms we have already discussed, so the decision meeting is about a known offer.
  7. Tripwire: IF the success evidence isn't met BY the end date, THEN we hold the decision meeting anyway and choose between converting, ending, or one extension on new written terms.

The last line keeps a missed milestone from turning into a silent extension. A pilot that ends on its date with a decision, even a no, is a better result for both companies than one that fades.

If you sell finished work rather than seats, the pilot is where the customer first sees that work done, which is the thread running through Software vs. Services and Outcome-Based Selling. For a raise, a converted pilot is evidence you can show, and my note on the ladder of proof covers how investors weigh that kind of evidence.

Frequently asked questions

What is a pilot for a startup?

A pilot is a short, defined period in which a customer integrates your product and tests what it does for them in real work. When the test is mainly about technical feasibility, people often call it a proof of concept.

How long should a paid pilot last?

Pepe Agell's March 19, 2026 guide for Pear VC says to aim for 1 month and keep a pilot under 3 months no matter what. Set the end date in the written pilot terms, and book the decision meeting before the pilot starts.

What should paid pilot success criteria include?

Write down the evidence both sides will accept: the integration is live, the customer can point to an early result in their own numbers, and, for usage-based pricing, you know what normal usage looks like. Name who signs off on each item.

Write the 4 agreements before the pilot starts. If you want help pricing the result you're piloting, get the free Sell the work skill, and for the rest of what I tell my founder friends, get the Sunday issue.

Get the next one.

One essay every Sunday on how to build leverage in your life and work.