The Refusal Test: one question that replaces frameworks, scoring models, and persona decks.

I have spent twenty years reading companies from the outside: customer language, financial filings, hiring patterns, what they fund and what they kill. It started with my agency and later at a Big Four consulting firm (where I worked side by side with firm leadership, sales coaches, bid/proposal teams, and SMEs to win the hearts and minds of buyers by helping them show up differently — over $2B won). Today, alongside my advisory practice, I’ve built a system that does this reading at scale. Across all of it, one pattern held: the work that actually changed a company was bought by one specific kind of person, and every ICP framework I have seen is engineered to find someone else.

Here is what kind of piece this is. It is not a segmentation exercise. It replaces your ICP document with one sentence and gives you one question to ask in the first meeting. Ten minutes to read. The cost is real: applied honestly, it will disqualify most of your pipeline. That is the point.

The sentence: my ICP is the person in an organization who has the authority to refuse revenue.

That is the whole framework. The question that puts it to work is called the Refusal Test, and the rest of this piece is why the sentence is correct and how to run the test.

I. What an ICP Is Actually For

An ICP exists to answer one question: where will what I sell actually produce the change I am selling?

Notice that is not the question the frameworks answer. The frameworks answer a cheaper one: who is most likely to say yes? Company size, industry, tech stack, funding stage, title keywords. All of it optimizes propensity to buy because closed revenue shows up this quarter, whereas realized value does not. So the machinery drifts toward whoever can approve a purchase fastest. For anyone selling advice, strategy, or intelligence, that person is usually someone who can sign but cannot act.

The milkshake study Clayton Christensen made famous is the cleanest demonstration that profiles are proxies. A fast-food chain segmented its milkshake buyers every demographic way it could, adjusted the product to match, and sales did not move. The variable that mattered was the job the milkshake was hired to do: a one-hand breakfast for a long, boring commute. The profile described the buyer. The job predicted the purchase. Most ICP documents have the same flaw: pages describing the company, with not a single line naming the variable that predicts whether the work lands.

The objection writes itself: that is a person, and an ICP is supposed to describe a company. Ask why. The account-versus-person split was never a principle. It was a proxy. Firmographics only ever existed to approximate a question nobody wrote down: can this organization act on what we sell? If you can test the question directly, the proxy is dead weight.

One honest boundary before we go further, and it cuts two ways. If your product pays off without the customer making any decisions (pure consumption, pure tooling), this test matters less to you; you have distribution math, not a positioning problem. And if you sell evidence rather than change, research, intelligence, ammunition for someone else’s decision, your buyer is often the person who needs to move a refusal rather than make one: the advisor arming a client, the operator arming a board. That is a legitimate ICP. It is a different sentence. The sentence in this piece belongs to anyone selling change itself, work that only pays off when the customer makes decisions that cost something.

Your first move: write down what the customer must do for your product to pay off. Not what they must believe. What they must do.

II. The Two Authorities

Organizations distribute the power to spend money widely and concentrate the power to refuse money at the top.

The mechanism is comp plans and P&L ownership. Everyone below the P&L owner is paid to increase revenue, so anyone who declines income is punished by their own incentive structure. That is why the authority to refuse pools where the P&L consequences land: founder, CEO, owner, sometimes a business-unit president.

Underwriters and credit officers look like the counterexample. They decline revenue all day. But ask who wrote the policy they decline inside. Refusing inside a policy is a job. Writing the policy is a position. Two different refusals, and only the second builds one: we do not serve that segment, we will not ship that feature, we walk from deals shaped like this.

And a committee is rarely what it looks like. Either it runs on veto, where any credible objection kills it, or it runs on deference, where one member’s judgment carries, and the rest ratify. In both cases, the refusal has a center of gravity: the member whose no would end it, usually the one who owns the P&L the decision lands on. That member is your ICP inside the committee. If you genuinely cannot find one, if the refusal has no owner anywhere, the organization cannot make positioning decisions at all. You have just learned it is not a buyer of change. At best, it buys evidence.

This is why “economic buyer” is the wrong test, and why my sentence is not BANT wearing a new coat. Budget authority is the power to spend money on you. Refusal authority is the power to act on what you sold them. Different powers. Frequently different people. A VP can hold a six-figure discretionary budget and zero ability to fire a customer segment. Every deal that closed with enthusiasm and died in implementation was a confusion of the two.

Volvo made this concrete in 1959. It gave away the three-point seatbelt patent, opening it to every carmaker and walking away from the licensing royalties it could have charged, on an invention since credited with saving more than a million lives. The sacrifice proved one concept: safety. No marketing department on earth could have authorized that decision. Only someone who owned the P&L consequences could. Every decision that ever built a position has the same signature.

And before you object that positions are also built through allocation, pricing, and hiring bars, look closer: every one of those is refused revenue in ledger form. Allocating capital to X is refusing the revenue Y would have produced. A high price refuses volume. A slow, expensive hiring bar refuses the revenue a faster hire would have booked. Opportunity cost is refusal wearing accounting clothes. The authority to refuse revenue is the authority behind it all.

Your second move: for each account you are chasing, write the name of the person who could make a Volvo-class decision there. That name is your ICP. Everyone else in the building is audience.

III. The Refusal Test

One question sorts every first meeting: “Tell me about the last time this company turned down revenue on purpose. What was it, and why?”

Watch the pronoun in the answer before you weigh the content. “I killed it” is authority. “Leadership decided” is proximity to authority. The pronoun tells you which one you are talking to faster than any org chart.

Then the answer itself lands in one of four places:

They have never refused, and could not. Wrong room. This person can champion you, and champions matter, but a champion is a route to the buyer, never the buyer. Stop pitching; start asking who could have made that call.

They have never refused, and could. This is your best customer. Not your easiest sale, your best one: all the authority, none of the position. Every dollar this company has accepted indiscriminately is a sentence in a positioning statement they never meant to write, and they are the one person who can start editing it.

They refuse, but the refusals do not rhyme. Authority exercised without a concept. Second-best customer. The muscle exists; the direction does not.

They refuse coherently, around one concept. Already positioned. This is a different conversation, defence and maintenance rather than construction, and if you are honest about that, they become your proof instead of your project.

The test qualifies and diagnoses in the same breath. That is what a real ICP definition does: it does not just tell you who to call; it tells you what the first conversation is about.

In-N-Out is the standing example of the fourth answer. Its refusal to franchise is decades old, expensive, and coherent around one concept, and it is the reason nobody needs the company to explain what it stands for.

This is the buyer-side half of an argument I make in Your P&L Is the Only Positioning Statement. That essay says the position lives in the pattern of refused revenue. This one adds the corollary: the only buyer who can change the position is the one who can refuse it.

IV. What This Deletes From Your ICP Document

The sentence does not eliminate your firmographics. It demotes them from definition to filters, and only three survive.

Reachability. Can you actually get to the refusal holder? This is why founder-led companies and businesses run by a P&L owner convert, and matrixed enterprises stall: because the authority sits five layers above your contact.

Stakes. Is the gap between what the company claims and where its money actually goes expensive enough to matter? Quaker paid $1.7 billion for Snapple and sold it for $300 million three years later, because what it thought it bought was not what customers were buying. Gaps priced at that scale fund any intervention. Small gap, small deal, no urgency.

Evidence. Is there enough public record to read them before the meeting? You cannot run the Refusal Test blind.

Everything else goes. Persona decks describing “the strategic, data-driven leader.” Lead scoring on title keywords. “Decision-maker” as a seniority tier, as if decisions were interchangeable. The one decision that matters has an owner, and your document should name the owner’s authority, not their demographic costume.

The exercise, if you want the diagnosis in numbers: pull your last ten deals that ended in “they loved it, and nothing happened.” For each one, write yes or no: could your contact have refused revenue? Count the yeses. That count is your ICP document’s report card.

Then rewrite the document.
One sentence, three filters. Done.

The Sentence, One Last Time

My ICP is the person in an organization who has the authority to refuse revenue.

Now notice what the sentence does. By construction, it refuses revenue. It rules out every buyer who could approve the work tomorrow and act on none of it. It shrinks the market for change to the people who can actually make it. That is how you know an ICP definition is real: it prices its own admission.

So run yours through the same test I have to run mine through. Write the definition. Then name what it costs you: which deals it forfeits, which pipeline it empties, which enthusiastic yes it tells you to walk past. If the answer is nothing, you have not written a definition. You have written a description of everyone who might say yes.

Write the sentence. Run the ten-deal count. And if the person you find in category two is you, all the authority and none of the position, that is exactly who the Clarity Kit was built for.

One sentence.
One question.
One authority.