Aizii Learn · AI Search

[← Back to Learn](https://aizii.ai/learn)

# How to benchmark AI visibility against competitors

A benchmark tells you whether your brand is easier to find than the brands a buyer compares you with, and where the gap is. It is not a score, because no vendor publishes a visibility score or a benchmark, and it is not a ranking, because no vendor publishes one. It is a comparison built the same way on both sides: the same prompts, the same assistants, the same schedule, the same log. A benchmark built any other way is a guess dressed up as a number.

Do not benchmark against every brand in the category. Benchmark against the three to five brands a buyer actually compares you with: the brands that appear in the same prompts, the brands your customers name when they ask for an alternative, and the brands that win the sales you lose. Three to five is enough. More than that is a list, not a benchmark.

## 1\. Pick the competitors the buyer picks

-   List the brands that appear in the same prompts as yours when you ask the assistants. Those are the brands the assistants already group with you, and they are the comparison that matters.
-   Add the brands your customers name when they ask for an alternative, even if the assistants do not name them yet. A customer who asks for an alternative to your brand is comparing you with that brand, whether the assistant knows it or not.
-   Add the brands that win the sales you lose, if you know them from returns, from support conversations, or from the market. A lost sale is a comparison with a result, and it belongs in the benchmark.
-   Keep the list to three to five brands. A longer list spreads the work thin and makes the comparison harder to read. A short list, tracked well, is a benchmark. A long list, tracked badly, is noise.

## 2\. Use the same prompt set for everyone

-   Use the prompt set from your own measurement: twenty to thirty buyer questions, covering your main products, your main categories, and your main buyer problems, plus a small set of comparison prompts that ask for the best option in a category or for a comparison between two product types.
-   Do not write prompts that favor your brand. A prompt that names your brand and asks whether it is good is not a benchmark prompt. A prompt that asks for the best trail shoe for wet weather, with no brand named, is a benchmark prompt, because every brand in the list has the same chance to be named.
-   Run the same prompts, in the same order, on the same assistants, for every brand in the list. Run the unbranded set first, so no brand name leaks into the unbranded answers. Log the same fields for every brand: named or not, recommended first or named among others, which pages were cited, and whether the facts stated were right or wrong.
-   Run the benchmark on the same schedule as your own tracking. Monthly is enough for most stores. Weekly is better when you are launching a product or watching a competitor closely. Pick one and keep it, because a benchmark run twice a year cannot show a trend.

## 3\. Compare the shares, not the anecdotes

-   For each assistant, compare the share of prompts where each brand was named. A brand named in eight of twenty prompts is more visible on that assistant than a brand named in three, and the gap is the benchmark.
-   Compare the share where each brand was recommended first, separately from the share where it was merely named. A brand named last among five is not the same as a brand recommended first, and the two shares can move in different directions.
-   Compare the share of mentions with wrong facts. A competitor with many mentions and many wrong facts is a competitor with a trust problem, and that problem is an opening, not a lead to chase.
-   Compare which pages get cited. If the assistant cites a competitor's product pages and your guide, the competitor's product pages are doing work yours are not. If it cites your product pages and a competitor's guide, the reverse is true. The page comparison shows where the gap is, not just that a gap exists.
-   Compare the trend, not the run. One run can be noise. Three runs in the same direction is a signal. A competitor whose share is rising while yours is flat is gaining on you, even if you are still ahead. A competitor whose share is falling while yours is flat is losing ground, even if they are still ahead.

## 4\. Compare the sites behind the shares

A benchmark that stops at the shares tells you who is ahead. A benchmark that compares the sites tells you why, and that is the part worth doing.

-   For each brand in the list, check the gate: is the store's robots.txt open to the AI crawlers, is the store password protected or behind a bot wall, and does the product sitemap exist and list the pages that matter. A competitor ahead on mentions with a closed gate is a competitor whose lead will not last, because the gate is the first thing to fail.
-   Check the index: are the competitor's key pages indexed in Google and in Bing. A competitor indexed in both, against a brand indexed in one, has a wider door, and the fix is a sitemap submission, not new content.
-   Check the facts: do the competitor's product pages name the price, the availability, and the shipping and returns terms in words a shopper can read, and does the Product structured data agree with the page. A competitor whose facts are complete and in agreement is a competitor whose pages are easy to cite.
-   Check the content: does the competitor have guides that answer buyer questions, category pages that say what the category is for, and an About page that says who the brand is. A competitor with guides that get cited is a competitor whose content is doing the work.
-   Write the comparison down, brand by brand, in the same fields. A comparison kept in memory cannot be acted on. A comparison written down, with the gap named, is a work list.

## 5\. What to do with the gap

-   When a competitor is ahead on mentions, do not copy their content. Find the gap the site comparison shows, and fix that gap on your own store, in your own words. A gap in the index is fixed with a sitemap and a recrawl. A gap in the facts is fixed on the page, in the markup, and in the feed. A gap in the content is fixed with a guide that answers a buyer question your store does not answer yet.
-   When a competitor is ahead on first recommendations, look at the pages the assistant cites for them. The cited page is the page doing the work. Compare it with your own page for the same product or category, and ask what it states that yours does not.
-   When a competitor is behind on facts, do not assume they will stay behind. A competitor with wrong facts today can fix them tomorrow, and the benchmark should be re-run after you fix your own gaps, not after you assume the competitor will not fix theirs.
-   When your brand is ahead, do not stop tracking. A lead in AI visibility is a lead in a system that changes every month, and a brand that stops tracking will not notice when the lead is gone.

### The fundamentals

These mechanics are shared across every engine and covered in full in their own guides:

-   •The llms.txt format and a complete sample, see our [guide to writing a good llms.txt file](https://aizii.ai/learn/ai-search/how-to-write-a-good-llms-txt-file)
-   •The markup this engine reads, see our [product schema guide](https://aizii.ai/learn/ai-search/how-to-add-product-schema-markup-for-ai-search)
-   •The complete crawler rule reference, see our [robots.txt guide for AI search](https://aizii.ai/learn/ai-search/how-to-write-robots-txt-for-ai-search)

### Say you run a coffee gear brand

Say you sell coffee gear, and you benchmark against three competitors: two larger brands and one smaller brand that wins the sales you lose. In the first run, the largest competitor is named in nine of twenty prompts on ChatGPT, you are named in four, the second competitor in five, and the smaller brand in two. On Perplexity, you are named in six, the largest in seven, and the smaller brand in five, citing its brewing guide twice. The site comparison shows the largest competitor's product pages state price, availability, and shipping in words and in markup, and its sitemap is submitted to both Google and Bing. Your sitemap was never submitted to Bing. The smaller brand's brewing guide is cited because it answers the buyer question your store does not answer. You fix the Bing submission, you write the brewing guide your store is missing, and you re-run the benchmark a month later. Your share rises on ChatGPT and on Perplexity, and the smaller brand's guide is no longer the only one cited. The benchmark did not tell you to copy anyone. It told you where the gap was, and the gap was yours to fix.

## What this guide does not cover

This guide covers how to benchmark AI visibility against competitors. It does not cover how to measure your own visibility in full, which is our guide to measuring your brand's AI visibility, or how to track mentions over time, which is our guide to tracking AI mentions of your brand over time, or how to run a full audit, which is our guide to running an AI visibility audit on your store. A benchmark tells you where you stand against the brands a buyer compares you with. The other guides tell you how to move.

## Check your own brand now

Run the free scan. It takes seconds, no signup, and shows whether AI crawlers can read your pages today: aizii.ai/aeo-scan

[Run the free scan](https://aizii.ai/aeo-audit)

## Sources

-   Google Search Central, Optimizing your website for generative AI features on Google Search (specific): [https://developers.google.com/search/docs/fundamentals/ai-optimization-guide](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide)
-   Google Search Central, AI features and your website (specific): [https://developers.google.com/search/docs/appearance/ai-features](https://developers.google.com/search/docs/appearance/ai-features)
-   Google Search Central, Creating helpful, reliable, people-first content (specific): [https://developers.google.com/search/docs/fundamentals/creating-helpful-content](https://developers.google.com/search/docs/fundamentals/creating-helpful-content)
-   Google Search Central, Product structured data (specific): [https://developers.google.com/search/docs/appearance/structured-data/product](https://developers.google.com/search/docs/appearance/structured-data/product)
-   OpenAI, Bots documentation (specific): [https://platform.openai.com/docs/bots](https://platform.openai.com/docs/bots)
-   Perplexity, Bots documentation (specific): [https://docs.perplexity.ai/guides/bots](https://docs.perplexity.ai/guides/bots)

Note: links marked specific point to the page that supports the claim in this guide. Links marked general docs point to the vendor's general docs section for that topic, not to a page that proves every step. Check the vendor docs for the current path before relying on a menu name or click path in this guide.
