Share of model is the percentage of AI-generated answers in a product category that name or cite a specific brand, measured by running a fixed set of buyer prompts across AI tools and counting appearances. It is the AI-answer counterpart to share of voice. To measure it, build a prompt list, run each prompt repeatedly in several tools, and divide your brand's mentions by all brand mentions.
Last updated: November 8, 2026.
Quick definition: what is share of model?
Share of model is the proportion of AI answers in a category that mention or cite a brand, compared with competitors, across a fixed prompt set and a fixed list of AI tools.
Summary
- Share of model counts how often AI answers name a brand against competitors for the same prompts.
- The core formula is the brand's mentions divided by all tracked brand mentions, times 100.
- Answer presence rate, mention share and citation share are three different numbers, so label each one.
- Results are sampled and vary by tool and by session, so run each prompt several times and keep the method fixed.
- There is no standard definition, so numbers from two tools or vendors are rarely comparable.
What is share of model?
Share of model is a marketing metric that measures how often a brand appears in answers generated by large language models, relative to its competitors. The unit of measurement is a brand mention or citation inside an AI answer, not a click, an impression or a backlink.
The metric is new and was coined by the industry, not by the AI companies. Marketing glossaries such as Anglera's glossary entry state that it has no standard definition and no first-party reporting behind it, so a number only means something when the method behind it is written down.
A category question such as "best payroll software for small business" shows the idea. Share of model tracks how often each payroll brand appears when buyers ask that question, and similar ones, in ChatGPT, Gemini, Perplexity and Google AI Overviews.
How does share of model differ from share of voice and rankings?
Share of model differs from share of voice because it counts presence inside AI answers, while share of voice counts presence in media coverage or advertising. It differs from a keyword ranking because an AI answer is sampled, so a brand can appear in most answers to one prompt and few answers to a similar one.
| Metric | What it counts | Where | Nature of the number |
|---|---|---|---|
| Share of model | Brand mentions or citations in AI answers, against competitors | AI assistants and AI search | Sampled and probabilistic |
| Share of voice | A brand's share of media coverage or advertising presence | News, social and paid channels | Counted from published items |
| Citation share | How often a brand's own domain is linked as a source | AI answers that show sources | Sampled and probabilistic |
| Keyword ranking | A page's position for a query | Traditional search results | A fixed position at a point in time |
A brand can have strong share of voice from earned media and weak share of model if its pages are hard for AI tools to retrieve or if the outlets AI tools cite for the category do not mention it. The two metrics should be read together. 5W's guide to AI share of voice and share of model compares the three metrics side by side, and the guide to tracking brand visibility in AI search shows how the numbers sit beside referral traffic.
How do you calculate share of model?
Calculate share of model by dividing your brand's mentions by the total mentions of all tracked brands across the same set of AI answers, then multiplying by 100. Pair that number with the answer presence rate, which is the share of answers that name your brand at all, because the two tell different stories.
| Number | Formula | What it tells you |
|---|---|---|
| Answer presence rate | Answers that name your brand ÷ total answers sampled × 100 | How often your brand shows up at all |
| Mention share | Your brand mentions ÷ mentions of all tracked brands × 100 | Your portion of the brands the answers name |
| Citation share | Answers that link your domain ÷ answers that show sources × 100 | How often your own pages are used as a source |
The table below uses made-up numbers to show how the first two formulas work. It is an illustration, not data about any brand.
| Illustrative input | Value |
|---|---|
| Answers sampled | 100 |
| Answers that name the brand | 40 |
| Total brand mentions across all tracked brands | 160 |
| Brand mentions | 40 |
| Answer presence rate | 40 ÷ 100 = 40% |
| Mention share | 40 ÷ 160 = 25% |
How do you measure share of model in five steps?
Measure share of model by defining the category and competitors, building a fixed prompt set, running each prompt several times in each AI tool, logging every mention and source, and calculating the three numbers on a schedule. Keeping the method identical each time is what makes the trend meaningful.
- Define the category and the competitor list, so every run counts the same brands.
- Build a prompt set from real buyer questions across the funnel, from category questions to comparison and pricing questions, and freeze the wording.
- Run each prompt several times in each tool, because the same prompt can return different answers from one run to the next.
- Log the date, tool, prompt, brands named, order of mention, sources cited and whether the description of your brand is accurate.
- Calculate answer presence rate, mention share and citation share, then compare with the baseline from the first run.
| Log field | Why it matters |
|---|---|
| Date and tool | Tools and models change, so results need a timestamp |
| Prompt wording | A small wording change can change which brands appear |
| Brands named and order | Feeds mention share and shows prominence |
| Sources cited | Shows which outlets and pages the tool relies on for the category |
| Accuracy of the description | An inaccurate description is a content problem to fix |
5W's AI visibility audit shows what a baseline report contains, and the GEO measurement guide explains which numbers deserve a place in monthly reporting.
Why do share of model numbers vary so much?
Share of model numbers vary because AI answers are sampled, tools update often, and vendors calculate the metric in different ways. Two tools run on the same brand can return different percentages from the same answers, so a number should always travel with its method.
- Sampling: each run is one draw from a range of possible answers, so small samples swing widely.
- Model updates: a tool update can shift results without any change in the brand's own activity.
- Formula differences: some vendors count mentions, others count answers that name the brand, and others count links.
- Prompt choice: a prompt set weighted toward one funnel stage favors some brands over others.
The practical rule is to write the definition down, record the formula, prompts, competitors and tools, and keep them fixed. Do not compare numbers across tools, and do not read share of model as a measure of revenue or clicks, since it counts presence in answers only.
How often should you measure share of model?
Measure share of model monthly with the same prompt set, so each result is comparable with the last. A quarterly cadence is the minimum, since models and sources shift often, and a monthly cadence catches changes within a reporting cycle.
Pair share of model with AI referral traffic from analytics. The first shows whether the brand appears in answers, and the second shows whether those appearances send anyone to the site.
What can move share of model?
Share of model moves when the pages AI tools retrieve become easier to quote and when independent outlets describe the brand the same way. A peer-reviewed study accepted to the ACM KDD 2024 conference found that adding statistics, quotations from relevant sources and citations to a page raised its visibility in generative engine responses by up to 40%, as described in GEO: Generative Engine Optimization.
The same paper reports that results varied by subject area. Google's guidance on AI features and your website says there are no additional requirements to appear in AI Overviews or AI Mode beyond the fundamentals of Search. The work that follows is page rebuilds, earned coverage and technical access, which the GEO optimization practice describes and the GEO services cost guide prices.
Frequently Asked Questions
What is share of model in simple terms?
Share of model is the percentage of AI answers in a category that name or cite a brand, compared with competitors, across a fixed set of prompts.
How do you measure share of model?
Run a fixed set of buyer prompts across several AI tools several times each, log the brands and sources in each answer, and divide your brand's mentions by all tracked brand mentions.
What is a good share of model?
There is no standard benchmark, because the metric has no standard definition and depends on the prompts, tools and competitors tracked. Compare a brand with its own baseline and with named competitors measured the same way.
Is share of model the same as share of voice?
No. Share of voice counts a brand's presence in media coverage or advertising, while share of model counts its presence inside AI-generated answers.
How is share of model different from citation share?
Share of model counts mentions of a brand in answers, while citation share counts how often the brand's own domain is linked as a source. A brand can be named without being linked.
How often should share of model be measured?
Monthly with the same prompt set is the working standard, and quarterly is the minimum, because AI tools and their sources change often.
Can share of model be guaranteed to rise?
No provider can guarantee a rise or a citation. Pages, earned coverage and technical access can improve the conditions for a mention, and the metric reports what changed.
What should a brand do first?
A brand should write down its category, competitor list and prompt set, run a baseline across at least three AI tools, and record the formula it will use every month. Teams that want the baseline built for them can start with 5W's AI visibility agency page, and the GEO audit checklist lists the checks that come with it.
