Product

Visibility Sentiment Competitors Citations AI Traffic Opportunities Pricing

Engine trackers

ChatGPT Google AI Overviews Perplexity Claude Gemini Grok Meta Spark AI Mode

Resources

How it works Blog Learn Log in
Recommendations

Other tools stop at the dashboard. Referenced hands you the fix.

After every scan, seven checks run over the new answers and turn what changed into work: the page to write, the thread to answer, the publication to pitch, the engine to go and win back.

Join the waitlist

7-day trial, no card

Action Summary · After scan 21 Aug 7 checks run
High

Write "choosing your first ultra shoe"

You are absent from nine answers that name two rivals. One page covers all nine.

High

Gemini is a blind spot

Your other engines average 58%; Gemini sits at 41%. Same prompts, different sources.

Medium

Answer the returns thread

Cited on five negative answers and still unanswered. Draft ready to edit.

Low

Pitch runnersworld.com

Cited on 21 answers, never beside you.

Low

Three untracked queries

Real impressions, no prompt covering them.

Healthy

Scans all complete

40 of 40 prompts answered on three engines.

Every line is computed first and written second. Nothing is posted without you. Refreshed every scan

Four kinds of opportunity

Prompts you are losing, questions you are absent from, threads to reply to, publications to pitch.

Drafts written for you

Replies and pitches arrive written, in your context, ready to edit.

Nothing auto-posted

No account connections to post on your behalf. You read it, you change it, you send it.

Refreshed after every scan

A recommendation appears when it becomes true and disappears when it stops being true.

What is GEO?

Generative engine optimisation is the work of getting named in the answer: being on the sources engines quote, answering plainly on a page they can use, and being described in terms a buyer wants to hear.

Not SEO with a new label

There is no ranking to climb and no page two to be relegated to. There is one answer, it names two or three brands, and you are in it or you are not.

SEO GEO
Rank in ten blue links Be one of three names
Optimise a page for a keyword Answer the question the way it is asked
Backlinks from anywhere Citations from the sources engines quote

What a recommendation is made of

Numbers first, sentence second. Every recommendation shows the facts it was computed from, so you can disagree with the advice without doubting the arithmetic.

Recommendation · 21 Aug High

Write "choosing your first ultra shoe"

Hoka and Merrell are named in all nine answers to this question across three engines. You are named in none of them, and no page of yours is cited on any of them.

Computed from Answers affected 9 Rivals named 2 Engines 3 Your citations 0
Facts come from the scan. Only the wording is written. See the nine answers →

Seven checks, run on every scan.

Each one is a rule with a threshold, not a prompt asking a model what it thinks. If a check finds nothing, nothing is written for it: a quiet week reads as quiet.

# Check
1

Content gaps

Prompts where a rival is named and you are not, weighted by rivals × answers.

Fires when

A prompt names a competitor in every answer and you in none

2

Visibility delta

This window against the last one, so nothing fires on noise.

Fires when

The score moves more than the usual daily swing

3

Engine blind spot

One engine well below the mean of your others on the same prompts.

Fires when

An engine trails your average by a clear margin

4

Sentiment cluster

A negative theme with enough classified mentions to be real.

Fires when

A criticism theme clears the five-mention floor

5

Unactioned discussions

Cited Reddit and Quora threads with no reply from you.

Fires when

A thread the engines quote is still unanswered

6

Untracked queries

Search Console queries no prompt of yours covers.

Fires when

Real impressions arrive on a question you never ask

7

Scan health

Failed engines, stalled runs, prompts that never answered.

Fires when

Any part of last night's scan did not complete

Four kinds of work, ranked by what it is worth.

Each kind comes out of the same scans, sorted by how many answers it would change, so the top of the list is always the cheapest win available to you.

Prompt opportunities · by deficit 14 OF 40
"Best cushioned shoe for a first 50k?" Hoka leads −100
"Most comfortable trail shoe for long days?" Hoka leads −67
"Trail shoes for wide feet?" Altra leads −34
Three of these are the same objection: cushioning over long distance.

Where dashboards stop.

Most tools in this category measure well and end there. The difference is not more charts; it is that the chart hands over a job with the reasoning attached.

Tells you the score moved
Names the prompt and the answer behind the move Sometimes
Ranks what to do next by answers affected —
Drafts the reply, the pitch and the page outline —
Shows the numbers each recommendation was computed from —

Recommendation questions, answered.

Where they come from, how often they change, and what stays your decision.

Does it post or publish anything for me? +

No. Every draft (a reply, a pitch, a page outline) is yours to read, edit and send. Nothing is posted automatically, and nothing is written to your site.

How often are recommendations refreshed? +

After every scan. A check that no longer finds a problem drops its recommendation on the next run, so the list only ever shows what is true right now.

Where do the recommendations come from? +

Seven checks run over your latest scan: content gaps, visibility swings, engine blind spots, sentiment clusters, unanswered discussions, untracked search queries and scan health. Each is a fixed rule with a threshold, not a model guessing, and a recommendation only appears when a check's threshold is crossed.

What counts as a content gap? +

A prompt where a competitor is named in every answer and you are named in none of them, weighted by how many rivals and how many answers are involved, not a single missed mention.

What do I get on the Brand plan? +

Three suggested actions a month, each with the numbers it was computed from and its reply, pitch or page outline drafted. Pro and Agency remove the monthly cap.

Stop reading dashboards. Start closing the gaps.

Free 7-day trial, no card: your first action list lands with your first scan.

Feature

Visibility

Your score across every AI engine.

Feature

Sentiment

What AI thinks of your brand.

Feature

Competitors

Share of voice, brand by brand.

Feature

Citations

Which pages AI actually cites.

Feature

AI Traffic

GA4 + Search Console, overlaid.

Feature

Opportunities

The fix for every gap, ranked.