The pilot goes well. Everybody is pleased. Your contact says some genuinely nice things.
Then you ask for the real engagement, and it turns out there is no budget line for it. Not because the work was bad, but because nobody ever had to create one, and creating one now requires a process that nobody has started and that will take two months you did not plan for.
That is how a successful free pilot fails, and it fails for an administrative reason rather than a commercial one.
Which is the argument for charging. Not much, and the amount matters far less than people think, but something, invoiced properly, with a purchase order behind it.
The paid pilot is about paperwork, not commitment
You will be told that charging creates psychological investment and that people value what they pay for. Possibly true, and it is not the real mechanism.
The real mechanism is that a paid pilot drags you through their procurement process while the stakes are low.
Think about what an invoice for four thousand dollars actually requires at a company of any size. Somebody raises a purchase order. You get set up as an approved vendor. Your bank details go in their system. Your insurance certificate gets filed. Your contract goes past whoever reviews contracts. Possibly a security review, a W-9, a diversity questionnaire, an onboarding portal nobody can log into.
All of that is one-time work. Do it at four thousand dollars and nobody looks closely. Defer it to the moment you are asking for sixty thousand and every one of those steps happens under scrutiny, in sequence, with your champion spending political capital on administrative friction rather than on the merits of your work.
A free pilot does not remove friction. It stockpiles it and hands you the bill at the worst possible moment.
That is also why the amount is not the point. The purpose of the fee is to exist, so the process runs. A pilot priced at a level that requires three approvals has defeated its own purpose.
Scope it to end on a decision, not a date
A pilot that ends on a date ends with somebody saying “great, let’s discuss next steps”, and next steps discussions have no owner.
A pilot that ends on a decision has the decision written into the scope before it starts. Three things, in writing.
One measurable outcome. Not three. The single thing that, if it moves, means this works. Pick something that can actually move inside the pilot window, which rules out most revenue metrics and is the reason pilots get scoped around leading indicators.
A named owner on their side. Not a department. A person whose name is on the document. This is also the qualifying question, and it is the one I would refuse to proceed without: if nobody’s name is attached, nobody’s reputation is involved, and nobody will fight for phase two in a budget meeting.
What happens on the last day, both ways. If the number is above the line, we proceed to phase two at a stated price. If it is below, we stop. Writing down the stop is what makes the proceed credible, and it is the clause buyers notice, because almost nobody offers it.
That third item is the same instinct as stating what is excluded, which I argued at length about proposals, and it works here for the same reason: a bounded promise reads as experience.
The fee that actually gets signed
Here is the structure I would use, and it is the one that survives a procurement conversation.
A pilot fee, credited in full against the first invoice of the engagement.
That sentence does more work than any discount. It converts the fee from a cost into a deposit, which changes which budget it comes out of and frequently which approval it needs. It is also honest: you do not want the pilot fee. You want the engagement, and you are willing to say so.
And it removes the objection you would otherwise get, which is that the buyer is paying twice to find out whether to buy.
What I would not do is discount the pilot heavily to win it. A pilot priced at a tenth of the real rate sets an anchor you then have to argue your way off, and the argument arrives at the exact moment you are asking for more money.
Two things to refuse
A pilot with no named owner. Covered above. It is the single highest-value refusal available to you and it costs you only pilots that were never going anywhere.
A pilot with no expiry on the price. This is how a pilot quietly becomes the engagement at pilot pricing: the work is going fine, the client is happy, nobody raises the next phase, and eight months later you are delivering a retainer’s worth of work at the introductory number. The fix is a date on the rate rather than only on the project. After the 31st, the price is the price.
That failure mode is a cousin of the one I described in retainer versus project pricing, where scope ratchets one way because adding takes an email and removing takes a conversation. A pilot with no price expiry is the same mechanism with a friendlier face.
The conversion numbers, and why I am not quoting them
You will find pilot conversion statistics and they are worth looking at once, for a reason the authors did not intend.
Published figures for corporate pilot-to-contract conversion run from about 6% to about 90%. Vendor analyses tend to put structured paid pilots at the top of that range and unstructured free ones near the middle or below.
I am not going to hand you a number from that, because almost all of it comes from companies selling pilot software or pilot consulting, and the sources that cite research firms mostly do not link to anything readable.
But sit with the spread. A range from 6% to 90% for the same nominal activity tells you that the pilot format is not the variable. What varies is who funded it, whose name was on it, and whether anybody wrote down what would count as success. Which is the whole argument of this post, arrived at accidentally by the people selling the opposite.
One piece of reasoning in that material is worth keeping, though. A paid pilot can convert at a lower rate than a free one and still be the better trade, because the buyer is covering your delivery cost. If free converts at 40% and paid converts at 30% but the buyer pays for the work, paid is usually ahead. Run that on your own numbers rather than taking anybody’s benchmark.
When a free pilot is genuinely right
Two situations, and I would not pretend otherwise.
Some buyers genuinely cannot procure anything without evidence. Public bodies, regulated industries, large enterprises with a mandated evaluation stage. There the free pilot is the price of entry, and the correct response is to make it small, time-boxed and documented rather than to fight it.
And if you are pre-product and the pilot is really your research, be honest about that with yourself. You are being paid in learning rather than money, which is a legitimate trade at a particular stage. Just do not scope it like a commercial engagement, and think about who owns what you build before you start rather than after.
The thing that does not work is the middle: a full engagement’s worth of work, given away, to a buyer who could have paid, with no named owner and no written definition of what would happen next.
If you’d like a second pair of eyes on a pilot scope before you send it, drop me a line on email, WhatsApp, phone or LinkedIn and we can have a quick chat, and my notes on getting marketing and sales pointed the same way cover what happens either side of it. I’m taking on a small number of contracts at the moment, so I’ll tell you straight whether it’s something I could help with.