Positioning

Positioning is a claim about a comparison someone is already making. The buyer will pick a frame of reference whether or not you choose one; the work is choosing it deliberately, naming the alternative honestly, and holding every differentiator to something the product does today.

When this fires

The product has to be described to people who do not already know it — homepage, deck, category page, launch, a new segment — or the existing description is not landing: prospects ask "so what is it", sales keeps losing to the same alternative, or two internal descriptions disagree. It does not fire for individual interface strings or for an ad headline; those are downstream of this.

Two traps this exists to catch

• Describing what you built rather than what it replaces. Feature lists are attributes. A buyer compares outcomes against the thing they do now.
• Positioning against the competitor you want. The real alternative is usually a spreadsheet, a manual process, an incumbent nobody loves, or doing nothing. Name the one they would actually fall back to this quarter.

Procedure

1. Pick one segment, and make it small enough to be wrong about. Who has this problem acutely, this quarter, with the authority to change it. "Teams that care about quality" is not a segment; "support leads at 50–200-person SaaS companies handling tickets in a shared inbox" is.
2. Name the trigger. What happens in that segment's week that makes them go looking. No trigger means no urgency, and no amount of wording fixes that — report it as the finding.
3. List the alternatives that segment would actually use, including the status quo and doing nothing. Source each from something real: lost-deal notes, support threads, review sites, what the user tells you. An alternatives list assembled from memory is a guess; label it one.
4. Inventory what the product actually does today. Read the product, not the roadmap. Keep shipped, in progress and planned in separate columns and never let the last two into a claim.
5. Reduce to the attributes only some alternatives have. Anything every alternative also has is table stakes — it belongs in the page, not in the differentiator. A useful test: would a competitor publish the opposite claim? If not, it distinguishes nothing.
6. Turn each remaining attribute into the value that segment feels, in a sentence they would say themselves. Take the vocabulary from their words — tickets, reviews, call notes, the phrases they search — not from internal naming. Quote the source for at least the central ones.
7. Choose the frame of reference: the category the buyer already understands and already budgets for, or the explicit "an alternative to X". A new category is a much larger bet than it looks, because you also have to teach the problem; take it only deliberately.
8. Write it as one falsifiable statement. For [segment] who [trigger], [product] is a [category] that [value] — unlike [alternative], because [proof]. If any slot is filled with something unfalsifiable ("best-in-class", "seamless"), the slot is empty.
9. Attach evidence to every claim. A benchmark that was actually run and its conditions, a named customer who said it, a capability you can demonstrate. A claim with nothing behind it gets cut, not softened.
10. Hold comparative claims to a higher bar. Naming a competitor and asserting something about them is a public, legally consequential statement. Verify it against their current documented behaviour, date the check, and route it for human review before it goes anywhere public.
11. List where this changes what is already published — homepage, pricing page, deck, app store listing, docs intro — as a diff of current wording against proposed. Publishing or editing any live public page is outward-facing: stop there and hand it over.

Checklist

• One segment named, specifically enough that it could be shown wrong
• The trigger written down, or its absence flagged
• Alternatives listed with a source, status quo included
• Every claim traced to shipped behaviour, with roadmap items excluded
• Table-stakes attributes separated from real differentiators
• Value stated in the segment's own words, with at least one quoted source
• Frame of reference chosen explicitly
• The statement written, with no unfalsifiable slot
• Comparative claims verified and dated, and routed for review
• Changes to live public pages proposed as a diff, not applied

Failure handling

• No segment can be named — say so. Undifferentiated positioning is usually a product or strategy question wearing a copy question's clothes, and wording will not resolve it.
• The only differentiator is price — that is a position, and a fragile one. Report it as the finding rather than dressing it as a capability.
• Every differentiator turns out to be table stakes — stop and say the honest result: the product's advantage is not visible from outside, or does not currently exist.
• The compelling claim is about something unshipped — it does not go in. Offer the shipped version of the claim, and flag the gap to whoever owns the roadmap.
• Two segments both look plausible — do not average them into one statement that fits neither. Write both, name what each costs, and let the owner choose.
• Evidence exists only as internal opinion — label it as an assumption to test, and say what would test it: lost-deal reasons, five customer conversations, a landing-page test.

Evidence to report

The positioning statement in full. The segment and the trigger, with what they came from. The alternatives list with sources. Each differentiator beside the shipped capability and the proof behind it. The buyer phrasing you adopted, quoted with its source. Which claims are assumptions rather than evidence, named as such. The proposed diffs to live pages, unapplied. And what you could not check — usually whether the segment agrees, which only they can answer.