Last verified: September 15, 2026
What You Can Now Do
Context Memo now tracks competitor features against your own build status in a single view, with a source link attached to every claim. You get one row per capability: what the competitor has published or announced, where that statement came from, and whether you've shipped the same thing. The row also shows how AI models describe that capability, so you can separate two problems that usually get confused. A real product gap is a capability a rival has and you don't. A visibility gap is a capability you shipped that models still credit to someone else. The task this changes: you stop rebuilding a competitive matrix before each planning meeting and start arriving with a gap list where every line traces back to a source anyone in the room can open.
Where It Is in Context Memo
This lives in the Competitor Intelligence area, alongside share of voice. The screen lists your tracked competitors with a row per capability. Each row carries three fields: the external evidence and its link, the internal status you set, and the date the claim was last verified. Rows sourced from model answers show which of the scanned models made the claim.
How to Use It
- Open Competitor Intelligence and select competitors you already track. Their published capability claims load as rows, each stamped with the source it was pulled from.
- Check the source on every row before you accept it. Open the link. If a claim rests on a prospect's secondhand comment in a deal note rather than the competitor's own documentation or changelog, mark the row unverified so it stops behaving like fact in your reporting.
- Set internal status per row: shipped, in backlog, or not planned. Assign the field to whoever owns the backlog. Status entered by the person who ran the scan is the fastest way to lose the room in a planning review.
- Filter to rows where the competitor is shipped and your status is not planned. That filtered list is your product gap set, with citations attached, ready to paste into a roadmap discussion.
- Filter the opposite direction: your status is shipped, but models don't name you for that capability. Those rows are positioning problems, not engineering problems. Generate a memo from the row to put the evidence on your own domain, then watch citation tracking for that capability.
- Turn on Slack alerts for new rows so the list moves between planning cycles instead of going stale until the next quarter.
Why We Built It
The complaint we heard repeatedly from product marketing: by the time the competitive matrix reaches the planning meeting, nobody trusts it. Someone built it once, left the team, and the doc kept circulating with rows nobody could source. The cost was concrete. Roadmap arguments got settled by whoever remembered a competitor's pricing page most confidently, and teams built features to close gaps that turned out to be a buyer's mistaken impression. The other half of the cost was quieter. Capabilities that were already shipped stayed invisible in AI answers, so competitors got cited for parity features and the loss never showed up in any report.
What It Does Not Do Yet
- No backlog sync. Internal status is set by hand. We don't read Jira, Linear, or your roadmap tool, so if a feature ships and nobody updates the row, the row is wrong. Assign the status field to one owner per product area and review it on the same cadence as your planning cycle.
- Only what's public or claimed. Rows come from competitor documentation, changelogs, pricing pages, and how models answer questions about that competitor. Unannounced work and private betas won't appear. Deal-sourced intelligence still needs a human to enter and verify it.
- Not a roadmap tool. Context Memo holds the gap list and the evidence behind it. Sequencing, estimates, and dependencies stay in the systems built for that work.
- Thin on new domains. If a competitor has little published material, or your own site has few pages for models to draw from, rows will be sparse and model coverage will be light. Publish foundational content on your domain first, then the gap view has something to measure against.