How to measure competitor presence in AI answers

A marketing manager asks this when a rival appears in ChatGPT answers and the manager has no data from that rival. This guide shows how one prompt set measures every brand at once, how to compare the rivals, and how to pick the list. The example is our own project, the one that what an AI visibility tool is describes. For the two share metrics and their denominators, read Share of Answer vs Share of Voice.

How do you measure competitor presence in AI answers?

You measure competitor presence in AI answers when you ask the questions of your buyers and count every brand each answer names. You need no data from the rival, because you count every brand in the same stored answers to the same questions. The same stored answers give the numbers of every brand, your own included.

The method has four steps:

1. Write the questions that your buyers ask. Leave your brand name out of them.
2. Ask each question on each AI platform every day. Store the answer text.
3. In each answer, mark each monitored brand that the text names. Count a brand once per answer.
4. Divide each brand's count by the same denominator. Then compare the brands.

Step 4 is why the method needs no rival data. Every brand faces the same questions, the same platforms and the same days. In the 28 days to 30 October 2026, on 7 AI platforms, the answers that named a monitored brand held 12,652 brand mentions. That is about 2.6 monitored brands per answer that named any. Our guide on what an AI visibility tool is gives the answer totals of the same window.

How do the rivals compare on one prompt set?

The rivals compare on visibility, Share of Voice and recommendation rate, all counted from the same answers. Our guide on Share of Answer vs Share of Voice defines the two share metrics, with a sheet formula for each.

The table holds the five rivals of our project in the 28 days to 30 October 2026, on all 7 platforms. Every number comes from the project's stored answers. Visibility divides by 8,953 answers. Share of Voice divides by 12,652 mentions. The last column divides by the brand's classified mentions: 1,670 of 3,270 for Otterly AI.

The three numbers rank the brands differently. Profound leads on visibility and on Share of Voice. Otterly AI leads on recommendation: 51% of its mentions recommend it, against 39% for Profound. The answers advised against a monitored brand in 74 of the 12,652 mentions. Our guide on AI visibility metrics says which decision each of these numbers supports.

The row of FullReach AI rests on 14 answers. 12 of the 14 mentions recommend the brand, a rate of 86%. One more answer moves a rate on 14 mentions by about 7 points. Read a rate beside its count.

Home›Competitors

Example

Share of Voice

HubSpot

31.0%

Pipedrive

22.0%

Salesforce

20.9%

Zoho CRM

14.6%

monday CRM

11.5%

How do you pick the competitor list?

Pick the rivals that your buyers compare you with, then check the list against the brands the answers name. A tool that reads every answer finds the brands that nobody monitors. In the 28 days to 30 October 2026, the project's stored answers named 1,708 brands that the project does not monitor. The most named of them, an SEO suite, appeared in 2,471 answers, 28%. That is nearly twice the answers of Scrunch.

Add such a brand only when your buyers compare you with it. Every monitored name takes part of the Share of Voice. With that SEO suite on the list, Profound's Share of Voice would fall from 29% to 24% on the same answers. Our guide on Share of Voice explains what else moves when the list changes.

Check the spellings too. A brand counts when the text contains its name or a spelling the tool knows for it. In the 28 days to 30 October 2026, one spelling of Otterly AI, with a space in the name, carries 3,333 of its 3,339 answers.

A first look at the answers to your own questions shows both the brands to add and the spellings to check.

Does the order of the rivals hold on every platform?

No. The order of the rivals changes from one platform to the next, so read the competitor numbers per platform. The table holds the visibility of each rival on four platforms, from the project's stored answers over the same 28 days. Each platform gave 1,303 to 1,307 answers.

Profound leads on all four platforms. Below it, the order moves. On Grok, Peec AI passes Otterly AI, and AthenaHQ passes Scrunch. On ChatGPT, Scrunch appears in nearly twice as many answers as AthenaHQ. On Perplexity, the two tie at 7%. The same five brands and the same questions give a different picture on each platform, and a pooled number hides that. Our guide on generative engine visibility gaps lists the prompts whose answers name a rival and not you, grouped by topic.

How do you read one rival head to head?

A head-to-head read narrows every number to your brand and one rival, so Share of Voice counts only the pair. FullReach AI returns this read under its Competitor filter, in the dashboard and through its MCP tools. The Head to head card on the Competitors page lists the pair platform by platform and topic by topic, with the gap in points.

In the 28 days to 30 October 2026, the project's stored answers hold 1,296 answers that named our brand or Scrunch. Scrunch holds 99% of the pair's mentions, and our brand holds 1%. The read gives one number per rival to report, and it shows where the gap is largest. On Grok, Scrunch appeared in 32% of the answers, and our brand in 1 of 1,306. On Perplexity, Scrunch appeared in 7%, and our brand in none of 1,303.

How often do you repeat the count?

Repeat the count every day, and report a window of weeks. In the 28 days to 30 October 2026, the project's stored answers put Profound's daily visibility between 37% and 47%. Its 28-day figure is 41%. A reader of one day could report either end. Share of Voice moved less: Profound's daily share ran from 27% to 31%.

Store every answer, because a stored corpus can be counted again with a new list or a new spelling. A count that only kept the totals cannot. Our AI ranking tool checklist gives twelve questions to ask a tool before you buy, cadence and data retention among them.

What to do next

1. Write ten questions your buyers ask, without your brand name. Ask them on one platform. Note every brand each answer names.
2. Pick three to five rivals from that list and from the names your sales calls hear.
3. Start a tool on the questions. After 28 days, read visibility, Share of Voice and recommendation per platform.

For how FullReach AI collects the answers, read the methodology.