Messaging & positioning

Positioning usually lives in a deck nobody re-checks. In KeystoneIQ it is one document, drafted from your evidence, tested against new evidence every week, and the same approved version your sales team, briefs, battlecards, and AI assistant use.

Messaging is the tab under Your product where your team keeps one living positioning framework per product: who it's for and not for, what buyers use today, why you win, how to talk about you against each competitor, the core message by channel, and your pricing stance. KeystoneIQ drafts it from your evidence, tests each "why us" claim against that evidence, and you steer what stays. Available on Growth and Pro.

Three ways to start

Start from	What happens
Draft from my evidence	KeystoneIQ reads your product's own pages, your documents, your team's answers, recent briefs, win/loss reasons, and competitor intel, then drafts a first version. The evidence card shows exactly what a draft will be built from and what is missing.
Paste what we say today	Paste your positioning or messaging doc. KeystoneIQ keeps your wording, structures it, and tests each claim against your evidence. Documents that are not positioning (pricing sheets, specs) are refused with a reason.
Answer five questions	Who buys, who should not, what they use today, why you win, what you must never claim (saved to Your stance, below). Takes minutes. Your answers are ground truth for every draft.

What the labels mean

Every claim under Why us carries a label and a hover with the source count and check date. Click the source count to open the pages and documents behind a Backed claim. Other sections (who it's for, alternatives, proof, per-competitor rows) are drafted from the same evidence but are not verified claim by claim; you own them by keeping or editing them.

Label	Meaning
Backed	Your evidence supports it. Capability claims are tested against your own pages and documents, contrast claims against that competitor's evidence, outcome claims against closed deals and calls.
No evidence yet	Nothing found for or against. The card says what would prove it.
Contradicted	Evidence says otherwise. The card shows the sentence when the check names where it looked.
Competitor says this too	The competitor claims it as well, so it stops differentiating.
Aspirational	You chose to keep it without proof.
Not checked yet	The check did not run; it retries on the next draft and on the weekly re-check.

Superlatives ("only", "best", "fastest") are never marked Backed; they need a third-party benchmark or customer quotes. Add them under Add proof and the claim counts as backed by you.

Steering

Every line has three verbs and, on claims, a fourth:

Keep: this is the team's call. A kept claim without evidence is served to sales and your AI assistant labelled Team's call, so a reader knows it is your word rather than proven fact. Keep is not offered on a contradicted claim; edit it, add proof, or keep it anyway (below).
Edit: your wording replaces the model's and carries into every future draft, word for word. The one-liner and elevator are editable in place too. Per-competitor cards are kept or left out as a whole rather than edited line by line.
Not this: pick why (Not true, Everyone says this, Not how we talk, True but don't lead with it, Legal won't allow it). Every reason does the same thing: the line leaves the framework and is never proposed again, nor are rewordings of it. You can bring it back later.
Add proof: a customer quote, a number you stand behind, or a link. The claim counts as backed by you.

Keep, internal only on a contradicted claim keeps it as aspirational: your team sees it, sales and your assistant do not, until evidence backs it (then it is served as backed). Imported documents start with every line kept, because they are your own words.

Your direction memo (what to lead with, what to avoid, who you are really for) and an optional paragraph in your voice shape every future draft. Redraft whenever you want; the receipt line says what the new version kept, left out, and learned.

Approve, share, export

Approve when it reads right. If a version still has contradicted or retire-suggested claims, KeystoneIQ asks first; approving anyway keeps those claims internal until you edit them or add proof. The approved version is what your team, sales, briefs, battlecards, and your AI assistant (the get_positioning MCP tool) see. Drafts are never served. Approval waits while a redraft is building. Up to six drafts an hour per workspace.
What readers see. One rule everywhere: a "why us" claim is served when it is Backed by evidence or explicitly the Team's call, and it carries that label on the share page, in exports, in the MCP tool, and in briefs. Kept-but-contradicted and aspirational claims stay internal. Proof points are served when the team supplied them, kept them, or they cite a stored source, with their type and date shown.
Share with sales creates a 90-day read-only page for people who never log in: one-liner, three things to lead with, why us (labelled), a card per competitor, quotable proof with its source type and date, and one-tap thumbs. Thumb counts show on each claim in Messaging. Revoke it any time.
Export Word and Export slides build a document or deck from the same rule. Tick Include unproven (marked) to add unproven lines, labelled.

It keeps itself honest

Every Tuesday approved claims are re-tested against the week's new evidence. A chip can change without a new version.
When a competitor now claims one of your differentiators or new evidence contradicts it, a positioning question appears on the Briefs page (the brief footer counts open questions). Answer it in one tap: keep it, retire it, or say you will reword it (a note lands on the claim in Messaging). Keeping a claim a competitor also makes keeps it in front of sales, labelled as your team's call; keeping a contradicted claim keeps it as aspirational, for your team only. A kept claim that becomes contradicted stops being served to readers until you act. Claims you edited are not re-judged weekly; your wording is tested on the next draft.
Messaging shows whether battlecards still reflect a brief from before the approval, and how much of the approved copy your own site actually says.
Briefs get a "your approved messaging" block their positioning and battlecard sections must use, and battlecards lead with the approved guidance for that competitor.

Your stance

Your stance is your team's rules for one product: what KeystoneIQ should never do or say, plus your take on the market. Messaging is how you want to be described; your stance is the short list of lines nobody should cross, wherever KeystoneIQ writes for you. For why written rules matter as AI drafts more of your competitive content, see AI messaging guardrails: what section D of the Frontier Playbook means for PMMs.

Where to set it

Settings → Your company & briefs → Your stance (the same tab holds brief context and, for owners, brief section order)
From any brief: the Your stance link at the end of the brief (Edit your stance, or Tell KIQ what it should never do before you have rules)
From Messaging → Your direction

What you can set

One-tap rules. Tick any of: Compete on price, Promise roadmap items, Name customers without approval, Use hype or #1 claims, Knock competitors.
Your own rules. Type what KIQ should never do, in plain words, up to 300 characters each.
Your take on the market. One or two sentences, up to 280 characters, used for framing. It is never cited as evidence.
Don't say this. Select a sentence in a brief and choose Don't say this. That sentence, and anything that makes the same claim, becomes a rule. Undo is right there. A flagged line also drops out of your battlecards (Growth and Pro) right away.

You can have up to 20 rules per product in total, counting one-tap rules, your own rules and Don't say this rules.

Your answer to "what you must never claim" in the five questions above is saved here too.

Where it applies

Briefs (weekly, deal, deep research, and pre-call briefs) are written to follow your stance. Weekly, deal and deep research briefs are then checked against your rules after they are written: a line that breaks a rule is rewritten or removed, and the brief's What we checked card says so, for example "Followed your 3 rules. Fixed 1 line (Compete on price)." Open Show what changed to see each line before and after. Shared links don't include the card.
Sales playbooks and board reports (Pro) are written to follow it too.
Ask KIQ (Growth and Pro) gets it as guidance when it frames an answer.
Messaging drafts treat your rules as hard constraints, and your take on the market frames the positioning.
Battlecards in the app follow it (Growth and Pro).
Your AI assistant gets it as team_rules from eight Context Layer tools, with an instruction to follow it and never do anything you listed. See the MCP reference.

Plans. Editing your stance needs Starter or above, and it is included in the 14-day trial. Rules you have already set keep applying on every plan, including Free. Workspace viewers can read the stance but not change it.

Brief context

Tell KIQ about your market once and every brief gets sharper. Brief context is five facts per product: your pricing model, your primary buyer, where you win, where you lose, and your stage, plus optional strategic priorities. Every brief uses them. Edit them from the Briefs page, the setup checklist on your dashboard, or Settings → Your company & briefs → Brief context.

Context Layer (MCP) reference: get_positioning serves the approved framework to Claude, Cursor, and other MCP clients.
Security: how workspace data, share links, and connectors are protected.