How Should You Structure a Paid Pilot?

How Should You Structure a Paid Pilot?

With a written success test agreed before it starts, a fixed end date, a named decision maker, and a price the buyer would be embarrassed to walk away from. A pilot without all four is a free trial with an invoice attached.

Paid pilots are the right answer for a lot of B2B companies. They qualify seriously, they fund the implementation work, and they give a cautious buyer a way to say yes without asking for a full budget.

They also go wrong in a consistent way, which is that they never end. Here is the structure we would use, and the two things nobody thinks about until later.

What Is a Paid Pilot Actually For?

Removing one specific doubt. Not demonstrating the product generally, and not building goodwill. If you cannot name the doubt in a sentence, you do not need a pilot, you need a better sales conversation.

Real doubts sound specific. Will this work with our data. Will our team actually use it. Will it hold up at our volume. Will it satisfy our compliance people.

Vague doubts sound like enthusiasm. "We want to see what it can do" is not a doubt, it is a request for a free consultant. That pilot has no finish line, because nothing can prove a vague thing.

How Do You Write a Success Test That Ends the Pilot?

As a statement both sides would answer identically without discussion. Write it down, put it in the agreement, and have the buyer's decision maker sign it rather than their champion.

The test needs three parts. A measurable outcome, a threshold, and a date. "Our support team resolves tier one tickets 20 percent faster by the end of week six, measured in your own reporting" passes. "The team finds it useful" does not.

Using their measurement, not yours, is the part that matters most. A result produced by your dashboard is a vendor claim. The same result in their system is evidence, and evidence is what survives the meeting you are not in.

The public sector has had a version of this discipline for years. The UK Government Service Manual's guidance on beta says teams should "start to gather data on how successful the service is based on the success metrics you identified in alpha," which is exactly the right sequence: the criteria come before the phase, not during it.

How Long Should It Run?

Long enough to produce the evidence, and not one day longer. For most B2B software that is four to eight weeks.

Shorter than four weeks and you are measuring novelty. Longer than eight and the pilot becomes the status quo, which is the failure mode to fear most. Nobody feels urgency about renewing something that already works well enough unsigned.

Put the end date in the contract as a hard stop rather than a review point. An extension should require someone to actively decide to extend, with a reason. Pilots that roll on quietly are the ones that never convert.

What Should You Charge?

Enough that it required a real approval, which is the entire point of charging. A price small enough to go on a card teaches you nothing about whether this company will buy.

Our usual shape is a fee in the range of one to three months of the eventual contract value, credited against the first year if they proceed. That structure does two useful things. It makes the pilot a genuine commitment, and it makes proceeding feel like the economically obvious choice.

What we would avoid is a discount framed as a pilot. Discounting to get someone into a trial sets your price anchor at the discounted number, and you will spend the negotiation climbing back, which is the trap we described in discounting to close a deal.

Who Needs to Be Named Before It Starts?

The person who will sign, the person who will use it, and the person on your side who is accountable. All three, in writing, before any work begins.

The missing name is almost always the signer. A pilot run entirely with a champion produces a happy champion and a decision that then has to be made from scratch by someone who saw none of it. That is where good pilots go to die.

So make the signer's involvement a condition. They do not need to attend every session. They need to have agreed the success test at the start and to have committed to a decision meeting at the end. If they will not do that, the deal is not real yet and the pilot will not make it real.

What Happens on the Last Day?

A meeting that was booked at the start, with only two possible outcomes. Proceed, or stop. Not "let us think about it," which is a third outcome you should explicitly close off.

Book that meeting in the first week, with the signer in the room. Send the results against the agreed test 48 hours before it, so nobody is reading numbers for the first time while deciding.

And genuinely allow the no. A pilot that cannot fail is not a pilot, and a buyer who feels they cannot say no will say "not yet" indefinitely instead. The clean no is a better outcome than the slow maybe, for both sides.

Is the Pilot Fee Revenue?

Ask your accountant before you invoice, because the answer affects how you report it and it is not obvious. This is the part that surprises founders.

IFRS 15 sets out the core principle as recognising revenue "to depict the transfer of promised goods or services to the customer in an amount that reflects the consideration to which the entity expects to be entitled." Its five step model asks you to identify the contract, identify the performance obligations in it, determine the transaction price, allocate that price across the obligations, and recognise revenue as each obligation is satisfied.

That sequence matters for a pilot because of the credit. If the fee is creditable against a future contract, you have to think about what the performance obligation actually is and how the price is allocated. A pilot fee that is really a prepayment for a later subscription is not the same as a fee for pilot services delivered.

None of this changes whether the pilot is a good idea. It changes what you can tell your board about it, which matters if pilot fees are a meaningful share of what you are reporting as revenue.

What Should the Agreement Actually Contain?

Less than a full contract and more than an email. Four or five pages that can be signed without legal spending a month on it.

The content we would insist on: the success test with its threshold and date, the hard end date, the fee and what happens to it on each outcome, who is named on both sides, what data you will access and what happens to it if the pilot ends, and what the price will be if they proceed.

That last item is the one most often left out, and leaving it out is expensive. A buyer who has proven the product and is then quoted for the first time has every incentive to negotiate hard, because you have already spent the leverage. Agree the onward price at the start, when the alternative is not doing the pilot at all.

When Is a Pilot the Wrong Move?

When the product is not ready, when you cannot afford the implementation effort, or when the buyer is using it to delay. Each is a real disqualifier and the third is the hardest to see.

A pilot requested with no signer, no named doubt and no date is usually procrastination with a process attached. The honest response is to say so, kindly, and offer a shorter and more pointed alternative such as a workshop or a structured reference call.

The other case to refuse is a pilot whose success test you cannot influence. If the outcome depends mostly on the buyer changing their own process, you will be blamed for a result you could not produce. That is a customer who needs a consultant first, and saying so is the same discipline as saying no to a customer who wants to buy.

What Would We Fix First in Most Pilot Programmes?

The written success test, and the decision meeting booked on day one. Those two changes fix more stalled pilots than anything else we could suggest, and they cost nothing.

After that, the onward price agreed up front, and a hard end date instead of a review date. Then charge enough that somebody had to ask permission.

The underlying idea is that a pilot is a decision process, not a demonstration. Everything above exists to make sure a decision actually gets made, by the person who can make it, on a date everyone already knows. That is unglamorous and it is the whole game, because the most expensive pilots are not the ones that fail. They are the ones that quietly never end. If you want to sanity check your own pilot structure before the next one starts, we are happy to look at it with you at phoenix.studio.