AI search visibility
Your clients are being recommendedin answers you cannot see.
An answer engine names three companies and moves on. There is no results page to check, no position to screenshot, and the same question returns something different tomorrow. Embrisk samples those answers continuously and turns them into a figure you can put in front of a client.
- Six engines
- 1,248 prompts sampled daily
- Observation counts on every figure
- ChatGPT+14.2%72%
- Google AI Overviews+8.6%61%
- Gemini+11.4%49%
- Perplexity−3.2%43%
- Claude+5.9%37%
- CopilotLimited coverage+2.4%29%
What changed
There is no page one in ChatGPT.
Your reports assume a ranked list and a position on it. An AI answer has neither. It names a few companies and moves on, so you need a different way to measure it.
| Dimension | Traditional search | Answer engines |
|---|---|---|
| The unit | WasA keyword | NowA question, asked in someone’s own words |
| The result | WasTen blue links, in a fixed order | NowOne synthesised answer, different every run |
| The win | WasRanking first | NowBeing named, and being the source that was cited |
| The evidence | WasA rank tracker | NowSampled answers, counted over a window |
| The report | WasPosition changed from 4 to 2 | NowNamed in 68% of answers, cited in 41%, up 12 points |
AI search
Know exactly how your clients appear in AI search.
Being mentioned is the least interesting thing an answer can tell you. What decides whether a client wins the category is where they land, which sources the model leaned on, and who it named instead.
- MentionsNamed at all
- PositionWhere in the answer
- CitationsLinked as a source
- SourcesWhich domains fed it
- CompetitorsWho else appears
- PromptsWhat triggered it
- Share of voicePortion of the category
- SentimentHow it is characterised
- ModelWhich engine and version
- GeographyWhere it was asked
- IntentWhat the asker wanted
Observations today
4,820
Prompts tracked
214
Engines
6
- 09:14“best accounting software for small businesses”ChatGPTCited#2
- 09:14“best accounting software for small businesses”PerplexityMentioned#5
- 09:13“alternatives to QuickBooks”AI OverviewsAbsent
- 09:13“payroll software for a 20-person company”GeminiCited#1
- 09:12“switching accounting platforms mid-year”ClaudeMentioned#3
- 09:12“accounting software with open banking”ChatGPTAbsent
- 09:11“best accounting software for startups”CopilotMentioned#4
Generative answers vary between runs. Every figure on this page is a sample estimate over a window, reported with the observation count behind it.
AI visibility
One client, every engine, one number you can defend.
Visibility is aggregated from thousands of sampled answers and broken back down by engine and competitor, so you can always say where a figure came from.
AI visibility
64.8%+18.4 pts
Across 6 engines, sampled daily on priority prompts
Visibility by engine
Share of sampled answers naming the client, with observation counts
- ChatGPT+14.2%72%
- Google AI Overviews+8.6%61%
- Gemini+11.4%49%
- Perplexity−3.2%43%
Share of voice
Category: small-business accounting
- Alder AccountingYour client+6.2%34.8%
- Ledgerly−1.8%27.1%
- Countway+2.6%21.4%
- Bookmark Books−4.1%16.7%
Position when named
Where the client lands in answers that mention it at all
- First named18%
- Second24%
- Third21%
- Fourth or later22%
- Not named15%
Coverage differs by engine. Where a provider limits what can be observed, the interface says so rather than filling the gap with an estimate.
Prompt intelligence
Track the questions that decide your client’s category.
Rankings used to be the unit of work. Now it is the question, and a client can lead one phrasing while disappearing from the next.
214 prompts tracked · refreshed daily
Demo dataSelect a prompt to see the competitor leading it, the engines it was observed on, and whether the client was cited rather than merely named. Prompts are grouped by intent so a research question is never scored against a buying one.
In practice
What an agency does with it.
Measurement is only worth paying for where it changes a decision. These are the three it changes most often.
Win the pitch
Sample a prospect’s category before the first call and show them where they stand against the three competitors they already worry about. A specific absence beats a generic audit.
Defend the retainer
When a client asks why AI mentions dropped, the answer is on screen with the prompts, engines and dates attached, not assembled over the following three days.
Price the new work
A measured gap is a scope. Prompt coverage, competitive position and citation share are all things a proposal can be written against.
Keep reading
Find out which of your clients AI is skipping.
Book 30 minutes. Name one client, and we will show you their real results on the call.
Name one client
Tell us which client you want to see.
We check it first
We sample their AI answers before the call, so the numbers are real.
30 minutes, live
See where they stand and what to fix first.
No slide deck and no contract on the call. If it fits, we run a guided pilot on one of your real clients before you pay for a full plan.