AI Product Marketing
Positioning statement

How to choose a market category for positioning

To choose a market category, pick it last: after you know what buyers do instead, the value only you give, and who cares most. Put the product in a candidate category, then list what a buyer would assume about price, buying process and who signs. In our method, wrong assumptions mean the wrong category.

Last updated by Omair Dawood.

On this page

The six moves of the category step

In our positioning method, naming the category is step 6 of nine. Steps 2 to 6 follow the method April Dunford sets out in her book Obviously Awesome (2019). In a post on the job of a market category, she writes that a category points a buyer toward your differentiated value. She also argues that starting positioning with the category gets the order wrong (Understanding the Job of a Market Category, read 2026-10-14). The moves below are our own words for step 6 and the record it leaves, in the order the method runs them.

  1. Finish steps 2 to 5 first. List what buyers do today, the attributes only you have, the value each one gives, and who cares a lot about that value. You cannot judge a category before you know the value it has to point at.
  2. Name one candidate category. The method defines the category as the frame that tells a buyer what to compare you with and what to expect. Write one candidate down, plus a runner-up, because the method's reasoning trail records both.
  3. List what a buyer would then assume. Put the product in the candidate and write three assumptions: what it costs, how it gets bought, and who signs.
  4. Check each assumption against the product. If those assumptions are wrong, the method calls the category wrong, however flattering it sounds. Try the runner-up. The method's record keeps a line for wrong assumptions and how the statement corrects them, so note those too.
  5. Escalate a category you have not claimed before. Our method refuses to decide this. Claiming a new category, or any category the brand has not claimed before, is a strategic call for a person. Draft the options, state the tradeoff, and hand it over as a question.
  6. Record the choice. The method's output keeps four lines for it. They are the category, what a buyer then assumes, which assumptions are wrong and how the statement corrects them, and whether it is a strategic decision. The reasoning trail adds one line on the runner-up and why it lost.

The Positioning statement tool runs this as its own stage, labelled "Picking the category", after it finds who cares most.

Worked sample: Miro's homepage

"Human collaboration at the speed of AI"

"Where your team and AI think, plan, and build together, in a workspace that works the way you do."

Read on . Source: Miro homepage

That is the H1 and the subhead we saw in a browser. Miro serves two subheads, and the one you get changes between loads. We quote the browser version. Most logged-out fetches the same day carried "The Intelligent Canvas where your team and agents think, plan, and build together" instead. Further down, a section heading and its second feature card say:

"Experience the Innovation Workspace"

"Intelligent Canvas"

"Empower teamwork on one, infinite, multiplayer canvas"

Read on . Source: Miro homepage

The browser tab title, which is not page copy, reads "AI Innovation Workspace | Miro".

What the copy does well. The H1 names the job, collaboration, before any feature. "Innovation Workspace" appears in a section heading and in the tab title, so one category name shows up in two places. In our reading, the canvas line is the most concrete on the page: "one, infinite, multiplayer canvas" describes something a visitor could picture and test.

What is missing, against the moves. Move 2 asks for one candidate. In our reading, the page offers three frames: "a workspace" in the hero, "Innovation Workspace" lower down, and "Intelligent Canvas". That last one is a feature name in one place and the frame in the other subhead. In our view, each frame sets up different comparisons. This homepage also cannot run move 1 for us. It does not show what buyers do instead, so moves 3 and 4 have no buyer evidence to check against.

The fix. First, a copy rewrite that commits to one of the page's own frames:

"The Innovation Workspace where your team and AI think, plan, and build together."

Then the category record, filled only from the page. Every slot the page cannot fill stays a question, because we will not guess Miro's buyers or what they assume.

Category             Innovation Workspace, the frame the section heading and tab title share (runner-up: Intelligent Canvas)
Buyer assumes        price [needs buyer evidence] · buying process [needs buyer evidence] · who signs [needs buyer evidence]
Wrong assumptions    [from buyer calls and lost deals: needs evidence]
Strategic decision?  [has the brand claimed this category before? a question for the owner]
Why runner-up lost   [needs the assumption test on both frames]

Moves 3 and 4 run on those open slots, so a team with call notes and lost deals can fill them. The page alone cannot.

Common mistakes

  • Starting with the category. The category is step 6 for a reason: it has to point at a value you have already found. Do steps 2 to 5 first.
  • Keeping a flattering category. The method's test is the buyer's assumptions, not how the name sounds. If price, process or signer comes out wrong, drop it.
  • Testing price only. The method lists three assumptions: price, buying process and who signs. So a category can pass on price and still fail on who signs.
  • Using several frames at once. In our view, two or three category names on one page leave the buyer to pick the comparison. Choose one and record the runner-up.
  • Claiming a new category quietly. Our method treats a category the brand has not claimed before as a decision for a person. Write it as an open question with both options and the tradeoff.
  • Forgetting the runner-up. The reasoning trail asks why the second choice lost. The method keeps that line so the pick can still be defended six months later.

Questions people ask

Should a startup create a new category?

Our method does not decide it: a new category goes to a person as a strategic question, with both options and the tradeoff. Dunford writes that about 10% of her clients are best positioned in a newly created category (her post, read 2026-10-14). She adds that most well-known tech products started in an underserved slice of an existing category. In our view, run the assumption test on an existing category first.

Should positioning start with the market category?

No. In our method the category is step 6, after alternatives, attributes, value and who cares most. Dunford argues the same in her post: you need the value first, or you have no way to judge whether a category does its job. A reader in the post's comments asked why the target market comes fourth in her order rather than first.

Sources

Pick your category with the Positioning statement tool

The tool asks four things: what your product does today, what buyers use instead, how you know they do, and what you say about yourselves now. It needs proof for at least two alternatives. You get a For, Who, Unlike statement, the five steps of April Dunford's method with reasons, what you should not claim yet, and questions only you can answer. It takes about 3 minutes of typing.

Write your positioning statement and pick its category