"Can We Start With a Small Pilot?" — When That's Progress and When It's a Trap

A buyer asking for a pilot is the most flattering way to lose two quarters I know.

It arrives sounding like momentum. They've seen the demo, they've asked the hard questions, and now they want to put your product in front of real users with real data. Nobody asks for a pilot on a product they hate. So the rep hears interest, says yes on the spot, and goes back to their manager with "they want to pilot it." Forecast moves to commit. Everyone feels good.

Then the pilot starts. IT needs three weeks for the security review. The one user who was excited goes on parental leave. Someone new joins the team and asks why they're evaluating this at all. The pilot's end date slips because "we didn't get a clean month of data." And when you finally circle back to talk about a contract, you're negotiating against a procurement team that now knows exactly how much you want this, and a champion who has already extracted most of the value he needed for free.

I've been on both sides of that. I've asked vendors for pilots I never intended to convert, when I was buying, because it was the cheapest way to get a problem partially solved without going through budget approval. And I've sold pilots that ate a quarter and produced a polite no. The difference between a pilot that closes and a pilot that eats your year is almost never the product. It's what you agreed to before the pilot started.

Three different things wearing the same word

When someone says "can we start with a small pilot," they are saying one of three things. They rarely tell you which. Your job on that call is to figure it out, because the correct response to each is different, and the response that works for one will kill the other two.

The de-risking pilot

This is the real one. The buyer wants to buy. They have a specific, nameable fear — the integration won't hold under their data volume, the accuracy will fall apart on their edge cases, their users won't adopt it — and they want evidence before they sign a number that will show up on someone's budget line.

You can spot it because the fear is specific. "We had a vendor last year whose API timed out every time we ran the nightly batch." "Our warehouse team abandoned the last three tools we bought them." A de-risking pilot has a shape. There is a thing that could go wrong, and the pilot is designed to find out whether it does.

These are worth running. They are also the easiest to convert, because the buyer has already decided to buy contingent on one variable. Resolve the variable and you have a deal.

The political pilot

Your champion believes. Your champion cannot get budget released. Maybe the money is locked until the next fiscal cycle, maybe there's a spending freeze above a threshold, maybe the VP who has to sign has never heard of you and won't take a meeting on a cold vendor.

So the champion asks for a pilot, because a pilot is small enough to sneak under a signature limit or fund out of a discretionary line, and because "we ran a pilot and here are the results" is a much easier internal pitch than "I read a case study." The pilot isn't an evaluation. It's ammunition. Your champion is going to use it to fight a budget battle you can't attend.

These are also worth running, but they need completely different handling. In a de-risking pilot you're proving the product works. In a political pilot you're producing an artifact your champion can carry into a room. Those require different success criteria, different reporting, and a different level of involvement from you during the pilot. If you run a political pilot like a technical one, you'll hand your champion a spreadsheet of latency measurements when what he needed was one slide with a dollar figure on it.

The stalling pilot

A committee cannot agree. Half the room wants to buy, half the room wants the incumbent, and one person wants to build it internally. Nobody wants to be the one who says no and owns that decision. So they say "let's start with a pilot," which is the corporate equivalent of "let's circle back." It defers the fight. It keeps everyone comfortable. It costs them nothing.

The tell is vagueness. Ask what a successful pilot looks like and you get "we'll see how it goes." Ask who's deciding and you get "we'll take it to the group." Ask what happens on the day the pilot ends and there's a pause, because nobody had thought about there being a day the pilot ends.

The stalling pilot is not a smaller version of a deal. It's a no that hasn't been said out loud yet, and every week you spend servicing it is a week you're not spending on a deal that could close. This is the one you have to be willing to decline.

The call where you find out which one you have

You don't diagnose this over email. You get on the phone and you ask four questions, in this order, and you shut up between them.

"What would you need to see at the end of it to move forward?"

Not "what do you want to test." What do you need to see to move forward. If the answer is a specific outcome — "we'd need to process a full month of claims without a manual correction" — you have a de-risking pilot. If the answer is "we'd want the team to get comfortable with it," you have a problem, because comfort is not a criterion anyone can fail or pass.

"Who else has to be convinced, and what would convince them?"

This surfaces the political pilot immediately. A champion running a budget play will usually tell you, sometimes with relief, because he's been carrying that problem alone. "Honestly, Dave won't approve anything he hasn't seen work here." Great. Now you know the pilot's real audience is Dave, and you can build for Dave.

"Walk me through what happens the day the pilot ends."

This is the one that breaks stalling pilots open. Real buyers have an answer. They've thought about the contract, the fiscal year, who signs. Stallers haven't, because in their mind the pilot doesn't end, it just fades. If you get silence here, ask it again more plainly: "Is there a meeting on the calendar after the pilot where a decision gets made? Who's in it?"

"If it works, what does the rollout look like and what's the budget for it?"

You are asking about money before the pilot. This feels aggressive. Do it anyway, and I'll explain why in a minute.

Run these four and you will know what you're dealing with before you've spent an hour of engineering time. Most reps skip them because the pilot request feels like good news and they're afraid of talking themselves out of it. You cannot talk yourself out of a real deal by asking a buyer what success looks like. You can only talk yourself out of a fake one, which is the point.

What you demand in exchange

A pilot is a trade. You're giving away product, support hours, and calendar. You're allowed to ask for things back. In my experience the reps who get burned are the ones who treat the pilot as a favor they're grateful to be granted rather than a mutual commitment with terms.

Four things. All four, in writing, before anything gets provisioned.

A named executive sponsor. Not your champion — someone above your champion who has been told this pilot exists and has agreed that a successful result means a purchase. You want that person on the kickoff call for ten minutes. If nobody above your champion will spend ten minutes on kickoff, nobody above your champion will spend six figures at the end. That's not a rule I invented; it's just how attention works in companies.

One metric, defined in their language. Not five. One. Written down as a sentence a stranger could evaluate: "By March 14, the system will have processed at least one full billing cycle with no manual reconciliation required." If you let them list eight criteria, you've given them eight chances to find a reason not to buy, and you've guaranteed a debate at the end about which ones mattered.

An end date on a calendar, with a decision meeting already booked. Not "about six weeks." A date. And a meeting invite for the week after it, with the exec sponsor on the invite, titled something honest like "Pilot results and go/no-go." Book it during kickoff, while everyone is enthusiastic. Getting a decision meeting on the calendar at the end of a pilot, when enthusiasm has cooled and the champion is dodging you, is one of the hardest things in this job. Getting it at the start is trivially easy. Do the easy version.

The price, agreed before the pilot starts. This is the one everybody skips and it's the one that matters most.

Why pricing has to happen first

Here is the leverage inversion nobody warns you about.

Before a pilot, you have something they want and haven't got. That's leverage. During the pilot, you're giving it to them. After a successful pilot, you have proven the product works, which feels like leverage and is actually the opposite — because now the only remaining question is price, and they know you have a quarter to close, a forecast with their name on it, and months of sunk effort you cannot recover by walking away. You have converted a two-variable negotiation into a one-variable negotiation, and you handed them the variable.

I have watched good reps win the technical evaluation cleanly and then get taken apart on the commercial one, and the pattern is always the same: the value was proven and the price wasn't fixed. That dynamic is bad enough in a normal deal — it's why the pricing conversation after a technical win needs its own preparation and its own script — and a pilot makes it worse, because the buyer now has usage data, internal familiarity, and a working deployment they could theoretically extend with a phone call. In regulated buying, where the security review itself can take a quarter, holding your number after the SOC says yes is a specific discipline for exactly this reason. The proof arrives, the leverage leaves.

So you fix the number first. The language is simple and you should say it without apologising:

"Happy to run this. One thing I want to handle now rather than at the end, because it's easier for both of us: let's agree the commercial terms for the full rollout before we start, contingent on the pilot hitting the criteria we've written down. That way, if it works, nobody's renegotiating in April — we just sign the thing we already agreed. And if it doesn't work, you've lost nothing."

That's a genuinely fair offer and it lands as one. What it also does is force a real conversation about budget, approval thresholds, and signing authority while you still have leverage — and it flushes out the stalling pilot immediately, because a committee avoiding a decision will not sign a contingent order form. They'll go quiet. That silence is information you'd otherwise have paid a quarter to buy.

Paid versus free

I don't have a study to hand you on paid pilot conversion rates, and I'm not going to make one up. But you don't need one, because the mechanism is obvious once you look at it.

A free pilot has no internal cost, which means it needs no internal approval, which means nobody's name is on it. Nobody's name on it means nobody is embarrassed when it dies. A free pilot can be abandoned quietly by simply not talking about it, and that is exactly what happens to most of them.

A paid pilot — even a small one, even one where the fee credits fully against the first year — requires a purchase order. A purchase order requires a signature. A signature requires someone to have looked at it and decided this is a reasonable use of money. That person now owns the outcome. They will chase the users who aren't logging in. They will show up to the results meeting. They have converted your deal from a vendor's problem into their problem, which is the single most useful thing that can happen in a sales cycle.

The fee is not about the revenue. Credit all of it back. The fee is about manufacturing an owner.

Declining a free pilot without losing the deal

Sometimes they won't pay and won't commit and won't name a sponsor. You have a stalling pilot. You need to decline it in a way that keeps the relationship intact, because stalling committees do sometimes turn into buyers later, when the pain gets worse or the political blockage clears.

Don't moralise. Don't explain your policy. Don't say "we don't do free pilots," which sounds like a rule and invites them to ask for an exception. Give them a reason rooted in their outcome:

"Let me be straight with you about why I'm hesitating. I've run these before without a decision date and a named owner, and they go the same way every time — everyone's busy, usage drifts, and three months later we're both a bit embarrassed and no closer to knowing whether this works. That's a worse outcome for you than not starting. So here's what I'd rather do. Give me thirty days, one team, one metric, and one person above you who's agreed that if we hit the metric we'll move forward. If you can't get that person to agree to it now, that's genuinely useful for me to know — it tells me this isn't the right quarter, and I'd rather come back when it is than burn your team's time."

Then stop talking. One of two things happens. Either your champion goes and gets the commitment, in which case you've upgraded a stalling pilot into a real one, or he tells you the truth about where this actually sits, in which case you've saved yourself a quarter and you can put a reminder in for next year.

The part reps get wrong is the tone. Said with an edge, it's an ultimatum and it kills the deal. Said as a genuine attempt to save both of you from a bad outcome — which it is — it usually raises your standing, because almost nobody talks to buyers like an adult with other options.

The political pilot needs a different build

If your diagnosis came back political, throw out the technical playbook. Your champion doesn't need a full evaluation. He needs a defensible number and a story.

Ask him directly: "What does Dave need to see on one slide?" Then design the pilot backwards from that slide. Fewer users, shorter window, narrower scope — whatever produces the cleanest version of that one number fastest. Then help him build the slide. Get on a call with him the day the pilot ends and write the internal pitch together, in his voice, with his framing, anticipating the objections you know are coming from the room. You've sat through the sceptical version of that room before; the same instinct that gets you through a demo where everyone is hunting for where it breaks is what you're lending your champion for a meeting you won't be in.

And ask what happens if Dave says no. Sometimes there's a second path — a different budget line, a departmental purchase, a smaller initial scope that clears a signature threshold. Champions often know these paths and don't mention them because they're focused on the front door.

What I'd do next

The reason most reps fold on this is not that they don't know the right answer. It's that the pilot request comes at the end of a good call, everyone's warm, and pushing back feels like breaking the mood. So the words don't come out. Knowing the script and saying the script under social pressure are different skills, and only one of them is built by reading.

If I were sharpening this, I'd take the four diagnostic questions and the decline language above and run them out loud until they sound like something you'd actually say rather than something you memorised — which is exactly the kind of low-stakes, high-repetition drilling we built DrillCall for, because the alternative is practising on live pipeline. Ten minutes on the pilot conversation before the call you know is coming is worth more than an hour of reading about it.

A pilot is not a stage in your funnel. It's a trade, and you're allowed to name your side of it. Buyers respect reps who do. The ones who don't respect it were never going to sign anyway.