Competitors: see the share gap and target the questions competitors win

What this page is for

Not just "a competitor showed up too", but which questions they win, where and why. The page compares prompt by prompt, gives the root cause and the fix, generates a content brief, and points you to re-tests to verify the result after publishing.

When to use it

• To see where you rank in AI answers and how far behind or ahead you are.
• To find a batch of questions where competitors appear and you don’t, and fix them together.
• To find questions you don’t track yet where AI recommends a competitor.

The page

The page header reads Competitor Intelligence. Below it: Scope (market, platform, and period: All time, Last 7 days, Last 30 days, Last 90 days), Config (Configure competitors, Export report) and Last scan. Data refreshes automatically after each scan.

Four tabs:

1. See the gap: five KPIs — Your AI appearance rate (with your rank among competitors), Leading prompts, Behind prompts, Citation gaps and Replacement alerts. Then the Share of Voice table: Brand, Appearance rate, Rec. share, Weighted rec. (each recommendation divided by the brands named in that answer: alone = 1, among 3 brands = 1/3), Avg. position and Sentiment. Hover a row and click Not a competitor to exclude a brand. Below are By engine and the Appearance-rate trend.
2. Prompt battlefield: one row per prompt with priority (P0 / P1 / Keep), the competitors named and the root cause (no comparison, no guide, no citable content asset and so on). Filter All, Strong, Weak or Missing. Rows you win have View evidence; the others have View countermeasure. Select several rows and use View countermeasure to fix them together.
3. Competitor-only sources: pages that cite a competitor but never you, with Draft outreach and View page.
4. Prompt gaps: questions you don’t track yet where AI recommends a competitor. Configure competitors first, then click Scan (each candidate is verified on 3 engines; usually 1–3 minutes). Results are saved; use Add to tracking or Add all to tracking (n), and Find another batch for more.

Common tasks

Configure competitors

Configure competitors → type a competitor name or domain and click Add, or pick from Suggested for you (AI-detected) → Save config. Share of voice, the battlefield and prompt gaps use this list first.

Fix a batch of prompts together

1. In Prompt battlefield, select 3 to 8 prompts with the same language and intent. If you mix languages or intents, the dialog suggests splitting them and offers keep only this group.
2. Click View countermeasure. The dialog lists the selected prompts and Articles AI is citing for this group. If some prompts already have an article, you can Keep only the ungenerated.
3. Generate brief, then Send to writer to generate the article; it goes to Content for review and publishing.

Verify after publishing

Seven days after publishing, the brief’s prompts are re-scanned automatically. For a controlled before / after comparison, create a re-test under Prompt Tracking → Retests and check whether your appearance rate rose there.

Common questions

The share table includes unrelated brands. Hover the row and click Not a competitor; it is excluded and the shares are recalculated.

How long does a prompt-gap scan take? Usually 1 to 3 minutes. Results are saved and shown immediately next time; Find another batch looks for more.

Where did the re-test section go? Re-tests now live in Prompt Tracking → Retests, where they compare target prompts with control prompts over the same period.

Why is an engine locked in the platform filter? Engines outside your plan are locked; the note points to the Pricing page.

If something goes wrong

• No data yet — run a scan in Prompt Tracking first.: the project has no results yet.
• No competitors configured yet.: click Configure competitors first.
• Scan failed — please try again later.: an engine timed out; click Scan again later.