Competitors

Use the Competitors screen to compare your visibility against the brands the AI engines name instead of you.

Competitors overview

A tracked competitor is measured alongside you on every scan: same prompts, same engines, same period. That’s what makes the comparison honest — it isn’t a scrape of their site, it’s what the engines said when asked your questions.

The screen has three parts: the tracked competitors and their numbers, the gaps against them, and an inbox of candidates you haven’t decided on yet.

Your plan sets how many you can track.

How to read the comparison table

Each row is a brand — you and each tracked competitor — with Visibility, Mention rate, Sentiment, Avg position and Δ vs you.

Δ vs you is the difference against your own visibility in the same period, in percentage points. It’s a snapshot comparison, not a trend.

Click a competitor to open its detail: trend against you, share of voice by topic, mention share by engine, how often you appear in the same answers, and the sources citing them.

How to see where they beat you

Where they beat you breaks the gap into three tabs:

• By topic — a matrix of you against the best competitor per topic.
• By sources — cited domains that verifiably mention them and none of whose cited pages mention you. These are publication targets, not trivia.
• By prompts — the prompts where they appear organically in the latest scan and you don’t.

Good to know When a window has few sampled answers, the screen marks it low data. Those shares are directional, not statistical — more prompts and more scans firm them up. Don’t rebuild a strategy on a low-data cell.

How to add a competitor

Click Add competitor and give the brand’s name and domain. A competitor without a website can be tracked by name alone.

Adding one doesn’t retroactively change past scans — it’s measured from the next one.

How to work the potential competitors inbox

Potential competitors collects everyone else on the field, merged from AI answers, market research and your onboarding audit. Each carries where it came from and whether it’s Confirmed by answers or still a research hypothesis.

Promote the ones that matter into your tracked set; dismiss the rest so they stop appearing. A dismissed candidate can be restored later.

This inbox is where a Context recommendation sends you when candidates have been sitting unreviewed.

What happens after a plan downgrade

Competitors over your new limit are Frozen: they keep their data and stay visible in the list, but they’re excluded from metrics, reports and probes.

Unfreeze the ones that matter — within your limit — or move to a higher plan. An unfrozen competitor is included again from the next scan.