Gumshoe AI Review 2026

Key takeaways

• Gumshoe's persona-first methodology is the real thing: it builds AI-inferred buyer personas with roles, goals, pain points, and decision context, then runs actual conversations against AI models as those personas. You see how recommendations change depending on who's asking. No competitor makes personas this central.
• It's a diagnosis tool, not an optimization platform. Compared with Promptwatch, Gumshoe has no AI crawler logs, no traffic attribution from AI platforms to your site, no Reddit or YouTube citation tracking, no ChatGPT Shopping insights, and content generation capped at 3 to 10 pieces per month with no CMS publishing.
• The "11 models" headline oversells the product. Paid plans run 6 models per audit and 4 per pulse audit. Tracking is English-only, and reviews flag no Microsoft Copilot coverage.
• Pricing is subscription-based per project (one brand, focus, location, and language per subscription): Starter $99/mo, Pro $299/mo with the full five-audit suite, Enterprise custom. The free sample audit needs no credit card.
• Best fit: brand and product marketing teams at mid-market companies that want persona-level intelligence on how AI talks about them, and agencies selling GEO audits with pitch-ready deliverables.

What is Gumshoe AI

Gumshoe is a persona-based AI search visibility platform from a Seattle startup, Gumshoe.AI, Inc., founded by Todd Sawicki (previously CEO of Zemanta) and Patrick O'Donnell (co-founder of Urbanspoon and MightyAI). The company announced a $2M pre-seed in May 2025 and now claims 10,000+ brands and agencies and 40,000+ reports generated. There are no verified G2 or Capterra reviews to cross-check those numbers against, for what it's worth.

The premise is one I find hard to argue with: the same question gets different answers depending on who asks it. A value-conscious newcomer shopping for running shoes sees a different brand list than a goal-oriented powerlifter. Most visibility tools scrape anonymous AI sessions, which tells you what AI says to a blank slate. Gumshoe runs conversations as specific buyer personas across ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Grok, and Google AI Overviews, and shows you what each kind of buyer actually sees.

The methodology holds up too. Gumshoe uses official model APIs rather than browser scraping, with engineered system messages that replicate logged-in user context (memory, custom instructions, conversation history), validated against the native interfaces. Full conversation transcripts are available for verification, which is more transparency than most tools in this category offer.

The five audits

The product is organized around five audits. Starter ($99/mo) gets the Visibility Audit plus technical audits; the Sentiment, Citation, and Content Audits ship only with Pro ($299/mo).

The flagship Visibility Audit builds AI-inferred personas per report, analyzes buyer criteria across purchase decision factors, and produces heatmaps by persona, topic, and competitor, plus a brand leaderboard tracking up to 30 competitors with per-model visibility scores and momentum indicators.

The Sentiment Audit runs direct brand-mention and head-to-head competitor prompts (which the Visibility Audit doesn't run), then sorts recurring themes into Pros, Cons, and Caveats, with a persona x topic heatmap scored -1 to +1. This was the most-cited gap in 2025 reviews of the tool, so it's good to see it shipped, even if it's locked behind Pro.

The Citation Audit fetches the actual pages AI models cite, extracts author, publisher, and publication date, and splits results into your citations, competitor citations, and third-party outreach targets. That outreach list is a genuinely useful starting point for digital PR planning.

The Content Audit crawls your site and groups buyer topics into "retrieval," "generation," and "defend position" issue types, with one-click content generation per topic. The Technical GEO Audit is a per-URL AI-readiness scorecard covering crawlability, structured data, page structure, and readability, with a prioritized fix list (JSON-LD, schema, metadata, content trapped behind JavaScript). Starter includes 10 technical runs per month, Pro includes 50, and recurring Pulse Audits (weekly on Starter, daily on Pro) keep data fresh between full audits.

Pricing

A few things worth knowing before you swipe a card. Each subscription is tied to a single project, defined as one brand plus focus plus location plus language. Track two products and you're buying two subscriptions. There's no published annual discount. And if you run into the "$0.10 per conversation" pay-as-you-go pricing in older third-party articles, ignore it: Gumshoe switched to this subscription model in 2026 and grandfathered legacy customers through the end of the year.

The unit economics deserve a hard look. Both paid tiers track the same 100 prompts (80 in the monthly audit, 20 in the pulse), so the extra $200 for Pro buys frequency and the three extra audits, not broader coverage. No tier sells more than 6 models per audit.

Strengths

• The persona lens is genuinely unique. This is the reason to look at Gumshoe, and it isn't marketing fluff. If you sell to clearly distinct buyer segments, say an apparel brand serving studio regulars and powerlifters, persona-level visibility tells you things a flat brand score cannot. Promptwatch supports custom personas too, but Gumshoe makes personas the entire engine.
• Data quality and transparency. Official APIs, a documented methodology, and full conversation transcripts for verification. Independent reviewers consistently rate its data rigor among the best in the category.
• Model coverage at entry price. Six models at $99/mo beats rivals that start you at three or four.
• Audit output is structured for action. Technical audits become prioritized fix lists; the Citation Audit hands you a ranked outreach list. Shipium's CMO reported a 557% increase in citations of their domain in under four months, and Hawke Media rebuilt client content structure around its findings.
• Agency-friendly. Free prospect audits for pitches, multi-client workspaces, role-based access, CSV/JSON exports, and a read API on the agency tier.

Limitations

• No AI crawler logs. Gumshoe gives you a visibility score but can't show you whether ChatGPTBot, ClaudeBot, or PerplexityBot actually crawled your pages, what they read, or what errors they hit. Promptwatch's Agent Analytics logs 400+ crawlers with a crawl-to-citation rate per page. You can't fix visibility you can't diagnose at the infrastructure level, and Gumshoe doesn't operate at that layer.
• No traffic attribution. Mentions never get connected to visits, leads, or revenue; you track AI referrals manually in GA4. Promptwatch's visitor analytics tie AI-platform traffic and conversions back to citations, which is the difference between reporting on mentions and reporting on business impact.
• Content execution is thin. Three generated pieces per month on Starter, ten on Pro, no CMS publishing, no scheduled workflows. Promptwatch's Content Agents plan, write, and publish to Webflow, Framer, or WordPress on a schedule you control, and its gap analysis is grounded in prompt volume and difficulty data so you know what's worth writing.
• No Reddit or YouTube citation tracking, and no ChatGPT Shopping or ads monitoring. Reddit threads and YouTube videos drive a large share of AI recommendations, and sponsored placements inside AI answers are the newest battleground. Promptwatch has dedicated reports for both.
• No prompt volumes, difficulty scores, or query fan-outs. Gumshoe generates persona prompts but doesn't tell you how often real people ask them or how hard they'll be to win.
• English-only, no Copilot. Each additional language is a separate paid project. Promptwatch monitors responses in any language from any country and covers Microsoft Copilot.
• The 11-model claim. The homepage says 11; paid plans deliver 6 per audit and 4 per pulse. Gumshoe's own comparison table discloses "6 per audit, from 11 tracked," which softens things, but the headline still sets the wrong expectation.
• No verified G2 or Capterra presence and no SOC 2. Fine for most teams; a real problem for enterprise procurement.

Who it's for

Gumshoe fits three profiles well. Brand and product marketing teams at mid-market B2B or e-commerce companies that need to know how AI positions them per buyer segment. Agencies selling GEO audits, especially those that want free prospect audits and heatmaps that make client conversations easier; Hawke Media's SVP says the visibility heatmaps win over skeptical clients. And in-house content teams that already have a production workflow and need sharp diagnosis, not another publishing tool.

If you're an enterprise running multi-market programs, need AI traffic tied to pipeline, or want monitoring to turn into published content without manual handoffs, the gaps above will bite quickly.

Bottom line

Gumshoe answers one question better than anyone else in the category: how does AI talk about my brand to different kinds of buyers? The persona methodology is the real deal, the audit suite is well built, and the free sample audit is worth running just to see your persona heatmap. But that's where it stops. No crawler logs, no traffic attribution, minimal content execution, English-only. It's a diagnostic tool in a market that is quickly moving to end-to-end optimization. If you want a platform whose agents diagnose why you're invisible and then plan, write, and publish the fixes, Promptwatch covers more models including Copilot, tracks Reddit, YouTube, and ChatGPT Shopping, attributes real AI traffic to your site, and backs it with 1,840+ customers and a 4.7/5 G2 rating, at a lower price at comparable tiers ($95 to $245/mo against Gumshoe's $99 to $299). Run the free Gumshoe audit for the persona view. Build your GEO program on the platform that closes the loop.

Promptwatch

Track and improve your AI search visibility

View more