"Can You Just Do a Free Pilot?" — Say Yes to the Right Version and No to the Rest

The first time someone asked me for a free pilot, I said yes in about four seconds. It felt like winning. They wanted the product badly enough to put people on it. I skipped straight past the part where nobody had agreed what "it worked" meant, and I spent the next eleven weeks doing unpaid implementation work for a company that eventually told me the timing wasn't right.

That pilot taught me the thing I now believe about every free pilot request: the words are always the same, but they are covering three completely different situations, and the response that saves one deal kills the other two.

So before you answer, you have to figure out which one you're holding.

The three things "can we do a free pilot?" actually means

One: genuine risk reduction

This is the good version. The buyer wants what you're selling. They believe you, mostly. But they have been burned by a tool that demoed beautifully and then died in their environment, or they have a technical unknown they can't resolve on a call — an integration nobody has tested, a data volume nobody has hit, a user population that has ignored the last three things IT rolled out.

The tell is specificity. A risk-reduction pilot request comes with a named worry attached. "We need to see it handle our ticket volume during month-end close." "I need to know our field techs will actually open it." "Our security team wants to watch the SSO flow before we commit." The buyer is not asking for time. They are asking for evidence about one thing.

These pilots are worth running, and often worth running free, because they are short, scoped, and end in a decision either way. The risk you are removing is real, and the deal is genuinely blocked behind it.

Two: the budget timing dodge

This one sounds identical on the call and is completely different underneath. The buyer wants the product. The buyer believes you. The buyer does not have money until the new fiscal year, or until the current vendor contract lapses, or until a headcount req gets converted, and rather than tell you that — because telling you that feels like losing leverage — they ask for a pilot.

The tell here is that the pilot request arrives after the technical conversation is basically done. Nobody has a real unknown. You've already answered the integration question. Security has already signed off. And now, suddenly, there is a need to "see it in action for a quarter."

The other tell is the timeline. Ask when they'd want the pilot to end and watch whether the date lands suspiciously close to the start of a budget cycle. If a buyer wants a ninety-day pilot ending December 31, you are not in a pilot conversation. You are in a procurement-timing conversation that has been dressed up as one.

This is fixable and it is often the best of the three, because the intent to buy is real. But you fix it by naming the budget problem, not by running the pilot.

Three: value extraction with no intent to sign

The third one is the pilot as a way to get the outcome without the purchase order. Sometimes it's malicious — a team that intends to use your tool for one project and then let it lapse. More often it's just organizational drift. A manager likes your product, has no authority to buy it, cannot get anyone senior to care, and a free pilot is the only version of "yes" available to them. They are not lying to you. They genuinely hope it turns into something. But there is no path from the pilot to a signature because nobody who can sign has been in the room.

The tell is the absence of an executive. Ask who would sign the order form if the pilot succeeded and listen for a name. If you get a role instead of a name — "that would go to the VP of Ops, I think" — you are holding version three. If you get "I'd have to find out," you are definitely holding version three.

The other tell is unlimited scope. Version-three pilots always want everybody on it. All the seats, all the regions, all the use cases, no end date, "we'll see how it goes." That is not a pilot. That is a deployment with the invoice removed.

Diagnosing it live, without sounding like you're interrogating anyone

You can usually sort all three with three questions, asked in a curious tone rather than a defensive one. The framing matters. You are not challenging the request. You are trying to design the pilot well, which is exactly true.

Start with: "Happy to talk about it. What would you need to see by the end of it to be confident this is the right call?"

A risk-reduction buyer answers immediately and concretely. A budget-timing buyer gives you something vague and pleasant because the real answer is "nothing, I'm already confident." A version-three buyer gives you something vague because they have never thought about it.

Then: "And if you saw that, what happens next? Walk me through what the buying process looks like from there."

This is the question that separates two from three. A budget-timing buyer will accidentally tell you the truth here — "then we'd put it in the FY26 plan" — and now you know exactly what you're dealing with and can stop guessing. A version-three buyer gets fuzzy, because there is no process on the other side of the pilot.

Finally: "Who else needs to agree the pilot succeeded? I want to make sure we're solving for their bar and not just ours."

If there's an executive, this surfaces them. If there isn't, the silence tells you everything.

I have run this sequence in one form or another for years and the thing I'd stress is the tone. None of it is a trap. You are asking questions any good vendor would ask, and the buyer who is genuinely trying to de-risk a purchase will be relieved that you're taking it seriously.

The terms that make a pilot worth running

When I do agree to a pilot, free or paid, it has four things in it. Missing any one of them and I have never seen it end cleanly.

A named success criterion, written down, in the buyer's words. Not "see if the team likes it." Something like: "Twenty-five field techs complete at least one job report per week in the app for four consecutive weeks." Or "the integration syncs opportunity data to Snowflake with no manual intervention for thirty days." The criterion has to be something that can be observed and disputed. If nobody could look at the result and say "no, that didn't happen," it isn't a criterion, it's a vibe.

Write it in their language. If you write it, they'll nod and forget it. If they write it — or you draft it and they edit it — they own it, and ownership is the entire point.

An end date. Not a duration, a date. "Six weeks" turns into eight turns into a quarter. "Ends March 14" is a thing on a calendar. And it should be short. Almost every pilot I have watched that ran longer than about a quarter of a normal sales cycle ended in nothing, because the urgency that started it had evaporated by the finish.

A signed order form contingent on the criteria. This is the one people flinch at, and it is the one that changes everything. The paperwork is done, negotiated, and signed before the pilot starts, with a clause saying it activates if the named criteria are met by the end date. If they aren't met, it voids. Nothing to sign later.

The reason this matters is that it moves all the friction to the front. Legal review, security review, procurement, the price conversation — all of it happens while the buyer is excited and motivated, instead of after the pilot when the champion has moved on and a procurement analyst who has never met you is asking why the line item exists. I have watched more pilots die in post-pilot paperwork than in the pilot itself.

It also tests intent beautifully. A buyer who intends to buy will find the contingent order form reasonable. A version-three buyer will not sign it, and you will have learned that in week one instead of week fourteen.

An executive who agreed to the criteria in writing. Not a forwarded email. Not "my VP is aware." A reply, from the executive, on the thread with the criteria in it, saying some version of yes, that's the bar. It can be four words. "Looks right to me." That is enough. What you need is a person with budget authority who is on record about what success means, so that when success happens you are not re-litigating the definition with someone who wasn't there.

The terms that guarantee it drags

The inverse list is short and you already know it. Unlimited users, because a pilot with everyone on it has no natural end and no control group and no story. No metric, because then the outcome is a feeling and feelings can always be overruled. No executive, because then success has no audience. And "we'll see how it goes," which is the phrase that has cost me more months than any other sentence in sales.

If someone insists on all four, they have told you what the pilot is. Believe them.

Converting a free pilot ask into a paid short-term deal

Here is the move I use most, and it works because it gives the buyer more of what they actually want rather than less.

When you diagnose version two — real intent, budget timing — don't argue about the pilot. Reframe to a short paid term.

"Totally reasonable, and I'd rather structure this so it actually works for you. When people ask for a pilot, what they usually want is a way out if it doesn't perform. A free pilot gives you that, but it also means we can't put a real implementation team on it, and honestly the version of this you'd be evaluating isn't the version you'd be buying. So let me offer something better: a three-month paid term at the same rate, with the success criteria we just talked about written into it. If we hit them, it rolls to the annual. If we don't, you walk and you've paid for three months instead of twelve. Same protection, real deployment, and I can get you resourced properly."

A few things about that language. It never says no. It grants that the underlying worry is legitimate, because it is. It puts the reason for the change on your side of the table — we can't resource a free pilot properly — which is true and doesn't make the buyer feel accused of anything. And it offers a concrete swap rather than a refusal.

The rate matters. Do not discount the short term to get it. A three-month term at the annual rate is normal. A three-month term at half price teaches the buyer that your pricing is soft, and you will pay for that lesson at renewal.

If they push back on paying anything at all, the follow-up is:

"Help me understand the blocker. Is it that you're not confident enough yet to spend anything, or is it that the money isn't available until the new year? Those are really different problems and I can probably help with the second one."

That question does more work than anything else in this post. Confidence problems get solved with references, a technical deep dive, or a genuinely scoped free proof of concept on the one unknown. Budget problems get solved with start dates, deferred billing, or a signed contract that begins next quarter — none of which require you to work for free.

What you must not do is give a free pilot to solve a budget problem. It doesn't solve it. In three months you'll be in the same place, except now your product is free in the buyer's mind and you have no leverage left.

Where the pilot comes back as a discount ask

Watch for the second act. A free pilot request that you converted to a paid term will very often reappear at the end as a price concession — "we took a risk on you, we'd like that reflected in the annual number." That's the same conversation with a different hat on, and the way you hold your number there is the same discipline as holding it after a technical win. If you sell software, the SaaS pricing negotiation script walks through the exact sequencing for that moment, including what to trade instead of price. In longer-cycle industrial deals the pilot-to-discount move tends to be dressed up as a site trial or a phased rollout, and the energy and utilities version handles the procurement dynamics that come with it. In life sciences it shows up as an extended validation period, which the pharma pricing negotiation script treats as the same structural problem it is.

The principle underneath all three: the pilot was the concession. You already gave one. If it converts to a discount ask, you are being asked to pay twice for the same thing, and you should say so plainly rather than splitting the difference.

When walking away is the higher-EV move

This is the part reps resist, so let me be direct about the math as I think about it.

A pilot costs you real time. Implementation hours, support hours, weekly check-ins, an internal champion inside your own company who has to care. That time has an alternative use, and the alternative is prospecting or working deals that already have a signature path.

So when you're holding a version-three pilot — no exec, no criteria, no order form, unlimited scope — running it is not neutral. It is a decision to spend a chunk of your quarter on a deal with no defined path to close, and it will feel like progress the entire time, which is what makes it dangerous. A stalled pilot is the most comfortable place in the pipeline. It has meetings. It has activity. It has a forecast line. It just doesn't have a buyer.

My rule is simple: if they won't sign a contingent order form and they won't put an executive on record about the success criteria, I don't run it. I say so kindly, I offer the alternative, and if the answer is still no, I go work on something with a spine.

The script for that:

"I don't think I can make this one work, and I'd rather tell you now than three months in. Without someone who can sign agreeing up front on what good looks like, what usually happens is the pilot goes fine and then nothing happens, and I've burned your team's time as well as mine. If you can get me thirty minutes with whoever owns the budget and we can agree on the bar together, I'll run it happily. If that's not available right now, let's park it and pick it up when the priority is higher."

Some of those buyers come back. A few of them come back with the executive, because the walkaway was the thing that made the request real internally. The rest were never going to buy, and you found out in one conversation instead of over a quarter.

What I'd do next

The hard part of all this isn't knowing it. It's saying it out loud, calmly, in the four seconds after a buyer you like asks you for something reasonable-sounding. That's a reflex, and reflexes come from repetition, not from reading. If you want to build the reflex, run the three diagnostic questions and the paid-conversion language against an AI buyer in DrillCall until the words come out without a wobble in your voice — the version where you sound relaxed is a completely different call from the version where you sound like you're defending your pricing, and the only way to find the difference is to hear yourself do both.

One last thing. The best outcome of a free pilot request is not a free pilot, and it's not a paid one either. It's the conversation the request forces you to have about what success actually means and who actually decides. Most deals never have that conversation. If the pilot ask is what gets you there, it was worth answering.