How Should You Price a Pilot Project in a Market You Have Never Sold Into?

What should I charge for a first project in a market I have never sold into?

Charge your normal rate for a deliberately smaller scope. The pilot should be cheap because it is small, not because it is discounted. The moment you lower the rate to win the first logo in a new market, you have set the price for every conversation that follows in that market.

I sell from Bengaluru to buyers who are mostly not in India, and every time I have moved into a new segment the temptation has been identical. The first buyer in an unfamiliar market feels harder to win, so the instinct is to make the number smaller. That instinct is the expensive one.

Here is how I think about sizing and pricing a first engagement, and the specific ways I have seen this go wrong.

What is a pilot actually for?

A pilot exists to reduce risk for both sides at a price where being wrong is survivable. The buyer is testing whether you deliver. You are testing whether this market's problems are the ones you are good at. Both tests need a real deliverable, which is why free pilots teach nobody anything.

The asymmetry people miss is that your test matters as much as theirs. Entering a new segment means you do not yet know what the work actually costs you there. Different buyer, different approval chain, different definition of done. A pilot is how you find out before committing to a price you will have to honour for a year.

That framing also changes what you put in the pilot. It should include the part of the work you are least sure about in this market, not the part you are most confident you can do well. Confident work makes a nice demo and teaches you nothing.

Why is discounting the wrong lever?

Because a discount changes the price while a smaller scope changes the purchase. If you sell your usual engagement at half price, the buyer now believes your work costs half price, and the second conversation starts from that number. Shrinking the scope keeps the rate intact and makes the first cheque easy to approve.

The other problem with discounting is that it selects for the wrong buyer. The people most attracted to a discount are the people most sensitive to price, and price-sensitive first customers in a new market will set the tone for your referrals there. You end up with a segment full of buyers who found you because you were cheap.

I work on fixed fees, and most of my projects land somewhere between one thousand and ten thousand dollars. When I enter something new, I do not move that band. I move where in the band the first project sits, and I cut the scope until the number is honest at my normal rate.

How small should a pilot be?

Small enough to finish in two to three weeks and to be approved without a committee. If the pilot needs procurement, legal review, and three sign-offs, it is not a pilot, it is a small version of the real sale with all of the same friction and none of the revenue.

The practical test is whether one person on the buyer's side can say yes. That constraint decides the price more than anything about your cost base, because it defines the ceiling below which a decision is easy. In most organisations that ceiling is well known internally and a direct question will get you the number.

Timeboxing matters as much as scoping. A pilot that drifts across two months stops being a test and becomes an unpaid relationship. I put a hard end date on it and I say what happens on that date, which is usually a decision meeting rather than an invoice.

What should the pilot price be anchored to?

Anchor it to the value of the decision it unlocks, not to the hours inside it. The buyer is not purchasing three weeks of your time. They are purchasing enough evidence to commit or walk away, and that evidence is worth a great deal more than the labour that produced it.

In practice I work backwards. What is the engagement this pilot leads to, what would that cost, and what fraction of it can a single decision-maker approve without escalating. The pilot price lives at the intersection of those two answers, and it is usually higher than a time-based estimate would suggest.

The second anchor is your own opportunity cost. A pilot displaces paying work, and in a new market it also displaces the learning you would have got from a familiar project. If the number does not cover both, the pilot is a marketing expense, and you should decide to spend it deliberately rather than discovering it later.

What do you give up when you price a pilot too low?

Attention, mostly. Work that cost almost nothing gets treated as though it cost almost nothing. The buyer schedules the kickoff loosely, sends the inputs late, and takes three weeks to review the output, because none of it is protecting a meaningful commitment.

You also give up the honest signal. A pilot that is easy to approve because it is small tells you something real about buyer intent. A pilot that is easy to approve because it is nearly free tells you nothing, since almost anyone will say yes to nearly free. The whole diagnostic value collapses.

The third thing you lose is the upgrade path. Moving from a heavily discounted pilot to full rate requires the buyer to accept a large increase, and the natural human response to a large increase is to renegotiate rather than to proceed. I have written about how I set a defensible number in the first place in how I price an AEO audit as a solo consultant.

How do you write the pilot so it converts into the real engagement?

Name the next step in the pilot document itself, with its price and its start date. The pilot should end with a decision that was defined before the work began, not with a proposal written after the fact when everyone's attention has moved on.

I write the scope of work so that the pilot deliverable is genuinely useful standing alone and obviously incomplete as a system. That is not a trick. It is an accurate description of what a two week engagement can produce. The buyer should finish it holding something they can use and seeing clearly what the rest would involve.

The document matters more here than in a normal engagement because there is no history to fall back on. I keep to the same structure I use everywhere, which I laid out in how I write a scope of work document, with one addition: an explicit line saying what the pilot is testing and how both sides will know the answer.

When should you refuse to do a pilot at all?

When the buyer wants the pilot to be free, when the scope they describe cannot produce a real result at any size, or when they cannot tell you what would make them proceed. The last one is the clearest signal. A buyer with no decision criteria is not evaluating you, they are stalling.

I also refuse when the pilot is obviously a way to get one deliverable cheaply with no intention of continuing. That is a legitimate thing for a buyer to want, and it is a legitimate thing for me to price at full rate as a standalone project rather than as a pilot. Naming it honestly usually ends the confusion in one conversation.

The harder refusal is the interesting buyer in a market I want, who wants terms I would resent later. I have learned to say no to those, because resentment shows up in the work and the reference you were chasing turns into a reference you would rather not have.

How do you know afterwards whether the pilot worked?

Ask two separate questions. Did the buyer proceed, and did you learn what the work actually costs you in this market. A pilot that converted but took three times the effort you expected has told you something important, and treating it as a win will price your next ten projects wrongly.

I track the second question deliberately because it is the one that silently sets your margin. A new segment often carries hidden work: unfamiliar compliance steps, a review culture you have not met before, time zones that halve the number of useful exchanges per week. None of that appears in the first estimate.

Selling across borders adds its own version of this, and I went through the specifics in selling from India to buyers in the United States. The short version is that the friction is real, it is learnable, and the pilot is the cheapest place to learn it.

What should you do next?

Take the new market you are trying to enter and write down the smallest deliverable that would genuinely prove something, to them and to you. Price that at your normal rate. If the resulting number is too large to approve easily, cut the scope again rather than cutting the rate.

Then write one sentence describing what the pilot is testing, and put it at the top of the document. If you cannot write that sentence clearly, the pilot is not designed yet, and no amount of careful pricing will rescue a test with no hypothesis.

If you are trying to price a first project in an unfamiliar market and the number feels arbitrary, tell me what you are selling and who you are selling it to. I will tell you how I would size it. Let's chat.