Competitive Benchmarking in AI Search for In-House Teams
In-house marketing and product teams need a repeatable way to measure their brand's AI search visibility against direct competitors, across models and buyer personas. This guide sets one up in Gumshoe, tracking up to 30 competitors and starting with a free sample audit.
Published Nov 2026
Roughly 49% of U.S. adults used AI chatbots in 2026, up from 33% in 2024, according to Pew Research. Your brand's presence inside AI answers is now a competitive variable you can measure. This guide walks through each step of a benchmarking workflow in Gumshoe and ends with a live competitive view across models, personas, topics and cited sources.
For why a competitor gets recommended over you in the first place, and the tactics to win those answers back, read How to See Why Competitors Win in AI Answers. This page covers the setup and the measurement.
1. Run your free sample audit to set a baseline
Gumshoe gives each person one free sample audit with no credit card required, so it is the place to start before you subscribe. The sample is a smaller Visibility Audit, 4 personas with 5 prompts each across 4 AI models, and it is enough to produce your first brand visibility score.
Gumshoe generates the setup for you, including suggested competitors. The sample cannot be edited or rerun, so treat it as a first read on what AI models say about you and which topics you already appear in. Every later competitive comparison is measured against this baseline, so treat it as a starting benchmark, not a final report.
2. Define your competitive set before you configure the audit
Before you adjust competitors in Gumshoe, write down the brands you intend to track. Gumshoe supports up to 30 competitors per project, so there is room for direct substitutes and the adjacent players that compete for the same AI recommendations. For lean teams, five to ten direct competitors give usable data without crowding the leaderboard.
Use your sales team's "lost to" data, your category's review site rankings and your own positioning to pick brands that buyers really consider alongside yours. Keeping the list outside the platform lets you track additions and removals over time, which helps when you explain leaderboard shifts to leadership.
3. Configure a Visibility Audit with your competitors
On a paid plan, set up your Visibility Audit and edit the suggested competitors to match your list. Build your own starts at $99/mo per project with unlimited seats, and each audit runs 8 personas and 480+ prompts across six AI models. See pricing.
Gumshoe tests each prompt as one of your buyer personas, as described in How Gumshoe Works, so the competitive data reflects how models answer buyers rather than generic keyword queries. Review the prompts so they mirror the questions your target personas ask while they consider and evaluate a purchase.
4. Read the Competitive Leaderboard to measure your standing
When your audit completes, open the Competitive Leaderboard. It ranks your brand against every tracked competitor and shows each brand's mentions, visibility, change since the last run, and a head-to-head comparison view. Use the competitor toggle and the "Only show competitors" filter to isolate your direct set.
Mentions tell you how often a brand appears across all audited conversations. Visibility is the share of answers that mention the brand, which makes it the more reliable figure for comparing brands. Record both after each run in a tracking sheet, so you can show trend lines to stakeholders without relying only on in-product history.
5. Interpret the per-model competitive breakdown
The leaderboard is an aggregate view, but competitive gaps often sit in specific AI models. Gumshoe tracks 12 model families across eight providers and treats Google AI Overviews as a separate surface from Gemini, so a competitor may dominate ChatGPT answers while your brand leads in Perplexity. See AI models Gumshoe tracks.
Use the Model × Competitor heatmap, one of four heatmaps in the Visibility Audit dashboard, to see which models drive a competitor's overall lead. This matters when you plan content or technical changes, because models weight different signals and a fix that helps in one may not carry over to another. Put the models where your target personas are most active first, rather than trying to close every gap at once.
6. Use the persona heatmaps to find specific gaps
Two more heatmaps, Persona × Topic and Persona × Competitor, show where your brand wins or loses with specific buyer profiles on specific subjects. A cell where a competitor scores high and you score low is a targeted gap: AI models surface that competitor over you when that persona asks about that topic.
For in-house teams, this turns benchmarking from a blended score into a prioritized list, because each gap connects to an audience segment and a content area your team owns. Gumshoe's research across 1,145 brands and more than 600,000 AI answers shows that visibility varies a lot by persona, so persona heatmaps surface what aggregate scores hide. See What the Average Hides.
7. Analyze cited sources to see why competitors rank higher
Visibility gaps between you and a competitor often come down, at least in part, to which sources AI models cite. The Citations section of your Visibility Audit lists every cited domain, with top citations, citation types, filters for category, model, persona and topic, and a CSV export. Click into a domain to see which prompts, personas and models drove each citation and which brands those answers mentioned.
If a competitor's visibility in a topic or model traces back to a publication, review site or other third-party page where you do not appear, that source is a concrete gap to close through outreach, contributed content or earned coverage. According to Gartner, 53% of consumers distrust AI-powered search results, so being cited by authoritative sources is also a trust signal, not only a visibility lever.
8. Add the Citation Audit for deeper source detail
The Citations section shows which domains are cited. The Citation Audit shows what is on those pages. It collects the URLs cited in audit answers, groups variations of the same link, fetches each page, and records its title, author, publication date and the brands that appear on it. For benchmarking, that means you can find an article that cites a competitor and read the framing, attributes and claims it makes.
The Citation Audit is included in the Playbook at $299/mo per project, or available as a $100/mo add-on on Build your own. The Playbook also includes the Sentiment Audit, which scores sentiment toward your brand from -1 to 1 by topic and persona. Teams running an ongoing competitive program save the manual work of opening every cited page themselves.
9. Run prompt experiments to build a competitive intelligence workflow
Testing prompts on purpose surfaces competitive intelligence that a default setup may miss. Try prompts that mimic how a buyer compares vendors, asks for recommendations in your category, or evaluates a specific use case, then compare the leaderboard and heatmaps across runs. Each full audit runs 480+ prompts across 8 personas, so you have room to cover your most important competitive scenarios.
Edits take effect at the next scheduled run. Write down which prompt structures consistently surface competitor mentions and which surface your brand. That record is the base of a repeatable intelligence workflow rather than a one-time check. To read where competitors win by persona and by cited source, see How to See Why Competitors Win in AI Answers.
10. Track competitive movement with scheduled audits
One audit gives a point-in-time measurement. A schedule gives the trend data you need to act on competitive shifts. Both paid plans include a monthly Visibility Audit and an automatic Pulse Audit, weekly on Build your own and daily on the Playbook. Each Pulse Audit runs 4 personas with 5 prompts each across 4 models, for 80 conversations.
Set a fixed review rhythm, such as the leaderboard after each monthly audit and the Pulse Audit data each week, so you catch competitive changes before they settle in.
11. Use the Content Audit to close the gaps you found
Once benchmarking shows the topics where rivals outperform you, the Content Audit turns those gaps into content priorities. It is included in the Playbook, or available as an add-on on Build your own. It shows each buyer topic with its visibility, sentiment, top competitor and a content opportunity rating of high, medium or low, based on low visibility, negative sentiment and whether a competitor is winning. It then recommends new pages to create and existing pages to improve, each with its evidence.
Gumshoe can generate FAQs, knowledge articles, competitor comparisons and how-to guides from those gaps. Build your own includes 3 pieces per billing period and the Playbook includes 10, with more at $25 a piece; the allowance resets each billing period.
Brittney Fred, Head of Search at Linkflow, put it this way: "It ties directly to the content we create and the links we build."
What you have at the end
Your team now has a live competitive benchmarking system. Let the next scheduled audit run after your changes, then compare the leaderboard movement against the content and citation work your team shipped in between.
- A Visibility Audit tracking up to 30 competitors you chose
- A Competitive Leaderboard showing where you stand on mentions and visibility
- A Model × Competitor heatmap showing where each model leans
- Persona and topic heatmaps mapping specific gaps
- Citation analysis explaining why those gaps exist, and a content pipeline for closing them
Benchmark your brand against your competitors
Start with a free sample audit, no credit card needed, then track your full competitive set on a paid plan.
Free to start · No credit card required