Minimum viable offer mistakes in the proof, price and exit

Every cut in a minimum viable offer is deliberate, and every mistake is a cut too. The quickest way to find the bad ones is to stop reading the offer as its maker and read it as the buyer who ends up disappointed.

A disappointed buyer says short, concrete sentences. Each section below starts from one and lists the cuts that make it available. Three cuts do the most damage on the page and after payment: the proof, the price and the way out. The promise and the delivery get a short section near the end.

What to take away

"I believed what the page said": the proof

Proof is everything the offer says about itself before payment. A first offer has the least behind it, which is exactly when the temptation to decorate it is strongest. Four habits turn a thin offer into a misleading one.

Claims added to make a thin offer look substantial. Time saved, results delivered, numbers served. A claim needs a reasonable basis before it runs, and a first offer is precisely when there is least behind any of them.

Friendly feedback presented as customer testimony. An early user's enthusiasm is real and it is not a customer endorsement, especially if they got the thing free or cheap. What has to be disclosed and what an endorsement has to reflect is set out in the FTC endorsement guidance.

Borrowing a competitor's numbers to describe your own results. Their figures came from their delivery to their customers. Reading their public material is worthwhile for other reasons, which is the subject of competitor validation, but nothing on their page describes what yours does.

Implying a customer base you do not have. Plural language, logos of organizations that trialled rather than bought, a wall of quotes from three people. The honest version, that you have three customers and this person would be the fourth, is a perfectly good sentence and survives contact with the truth.

The table pairs each claim with what has to sit behind it before the page goes live.

"What was I charged?": the price

A buyer rarely complains about a price they were shown plainly. The complaints come from prices that were hidden, discounted or shaped to suit the seller's test rather than the buyer's decision.

Leaving the price out to remove friction. An offer given away tests interest and nothing else. The refusal is the evidence, and there is no refusal without a number. The ways this goes wrong are collected in pricing validation mistakes.

Pricing at the floor to make the test easier. You then learn that people will accept a price at which you cannot operate, which you knew. The floor includes delivery, support, rework, and the cost of getting one customer through a repeatable route, and only recurring costs bear on whether it works at volume.

Discounting to get the first yes. The first yes at a discount is not a data point about the price you intend to charge, and the customer who received it becomes a reference at the wrong number.

Charging monthly for something delivered once. The structure question is separate from the level question, and running them together makes both unreadable. That separation is the substance of pricing validation.

"I could not cancel": the way out

This is the last thing to cut, and usually the thing that makes a small offer defensible. Three mistakes remove it.

No refund, because the offer is cheap. The cheapness is the reason it can be refunded easily. A refund on a small first offer costs you very little and buys the right to sell something unfinished. The rest of this pattern is covered in offer testing mistakes.

A stated delivery date nobody confirmed. If money changes hands before delivery, the window you state is an obligation. For goods sold by mail, online or phone, the FTC's order rule says that when you cancel instead of delaying, you notify the customer and send the refund within the time a delay notice would have been due.

A subscription with an awkward exit. Where billing recurs, the buyer's silence becomes consent to the next charge, which the FTC describes as marketing in which the absence of affirmative consumer action constitutes consent to be charged.

An awkward exit also removes your best evidence, since the people who leave are the ones who can tell you what the offer was missing.

The promise and the delivery, in brief

Two cuts sit outside the three above. The promise cannot be minimum: narrowing it after the sale is a refund, not a clarification. A buyer who paid for X and got Y has been sold something else.

Delivery by hand is a legitimate cut, and hiding it is the mistake. If the customer believes software produces their result and a person does, they will discover it, and the discovery costs more than the disclosure would have.

Cap how many you can serve, say why, and stop selling when you reach it. A real cap is also the only honest form of scarcity.

The one question that catches most of them

Before shipping, write the sentence a disappointed customer would say about this offer, in their words, and then read it out loud.

"I paid for X and got Y"; "I thought a person would look at it"; "I could not tell it was you doing this by hand"; "I could not cancel."

If any of those sentences is available to a reasonable person, undo the cut that made it available. It will be cheaper to undo now than after the first sale.

Common questions

Is it dishonest to sell something not built yet?

No, if the customer knows what they are buying and when it arrives, and if you can honor both. It becomes dishonest at the point where the description implies a thing that does not exist.

How small can the first offer be?

Small enough that you can deliver it well ten times, and no smaller than the promise. Those two constraints usually meet in one place.

Should I say the delivery is manual?

Yes, plainly, and usually as an advantage. "A person checks every one of these" is a stronger sentence than most of what you would put in its place.

What if I already sold something I cannot deliver?

Tell every buyer now, offer the money back, and keep whoever wants to wait. The version where they find out costs several times more.