You have described the company four different ways this month. The deck says one thing, the site says another, and on a sales call you reach for whichever version the person in front of you seems most likely to understand. Every version is true. None of them was decided.
Positioning is four decisions, and a default is still an answer.
Founders treat positioning as a wording problem, so they hand it to a copywriter. It is four decisions, and your company is already running on an answer to each of them, chosen or not.
Who it is for. Not the total market you would take money from, but the specific person whose situation you can describe well enough that they recognize themselves in it.
What it replaces. Every purchase displaces something: a competitor, a spreadsheet, an agency, a person doing it by hand, or nothing at all. Whatever it replaces sets the budget line, the first objection, and the comparison the buyer runs without telling you.
What it is not. The uses you turn down, the buyers you send elsewhere, the features you will not build. This is the decision that makes the other three believable.
The proof. The reason a stranger should accept the first three from you rather than from anyone else saying something similar.
Leave any of the four open, and whoever is talking fills it in. Sales fills in the buyer. The loudest prospect fills in what it replaces. A competitor’s site fills in what you are not.
The same product can be positioned two ways.
Take a product invented for this example. It is not a real company and not a client: B2B software that watches a company’s nightly data jobs and flags the ones that finished successfully while producing wrong numbers.
Position it the first way, and it is a data quality platform for modern data teams. Who it is for: data teams. What it replaces: unstated. What it is not: unstated. The proof is a feature list.
Position it the second way: it is the thing that stops the finance team from opening Monday’s report and finding last week’s revenue restated. It is for the analytics engineer at a company where the finance close runs off the warehouse. It replaces the handwritten checks that person wrote after the last incident, and the hour every Monday spent confirming it has not happened again. It is not an observability suite, and it is not a reporting tool. The proof is a broken join caught before the report went out, with the timestamp.
Same software. Everything downstream of it changes.
Both versions send you a bill.
The first version is cheap to say and expensive to run. Every sales call starts with teaching the category, because nobody wakes up looking for a data quality platform. The roadmap gets pulled apart by four prospects, each meaning something different by the phrase. Nothing compounds, because no one searches for it, no room gathers those people, and no sentence a customer can repeat to a colleague. The spend goes into explaining instead of into reaching.
The second version is expensive to say and cheap to run. It costs you every company whose finance close does not touch the warehouse, which is most of them. It costs you the deals where the buyer turns out to be someone senior with a different problem. It commits the roadmap to one person’s Monday, and you can be wrong about that person. What it buys is a buyer who recognizes themselves in the first line, a comparison you chose rather than one you inherited, and a sentence that survives being repeated to whoever signs.
Both versions cost something. Only one of them is a decision. The cost of the first one arrives late: a year on, nothing built on top of it works, and none of it tells you why.
The fourth decision is the one that makes the sentence true.
Who it is for, what it replaces, and what it is not can all be written in an afternoon. Proof is the part you cannot write. You either have it, or you go and get it, and that is the whole difference between a position and a wish.
Proof is not a wall of testimonials, and it is not a claim with an adjective in front of it. It is what a skeptical buyer would accept without you in the room: a demonstration they can run themselves, a customer who will take the call, a public artifact that would be embarrassing to publish if it were not true.
There’s a science to product-market fit, and this is why Positioning sits second in what we do, after Brand and before Website. Brand decides what the company looks and sounds like. Positioning decides what it claims and to whom. The site is where the claim gets made in public, and it cannot be written honestly until the claim is settled.
Most repositioning work that reaches us never had a wording problem. Just four decisions nobody made, arriving late, with a year of copy, decks, ad accounts, and assumptions built on top of the gap.
Your sentence is decided when other people can repeat it.
There is a cheap test. Ask three people on the team, separately and in writing, who the product is for and what it replaces. Ask a customer the same two questions. If the answers differ, the decision has not been made. It has been delegated, and it is being made again every week by whoever happens to be talking.
The cost of leaving it open is easy to spot once you look for it. A long sales cycle with a lot of education. A roadmap that cannot say no. Channels that do not work, in any order. A brand carrying meaning the words never supplied.
We would rather spend the first week of an engagement forcing four decisions into one sentence, with proof named beside it, than write copy driven by answers no one has provided. The sentence is short. Earning it is the work.
Tell us what you’re building.