Mentions.so Review 2026

Key takeaways

• Mentions.so is a lightweight AI visibility tracker from the SEO agency Contact Studios. Plans run from $49 to $399 per month, with unlimited seats on every tier.
• It covers the basics well: visibility score, position, sentiment, share of voice, competitor ranking, source domains, and a live feed of model answers.
• It is a monitoring-first product. Compared with Promptwatch, it lacks IP-verified AI crawler logs, a content generation agent with CMS publishing, prompt volume and difficulty scoring, query fan-outs, Reddit/YouTube/LinkedIn citation reports, and AI Shopping and Ads tracking.
• The Agency plan ($399/month, white-label, pitch workspace, unlimited sites) is the strongest part of the offer.
• Users supply their own prompts, and the vendor's testimonials and "100+ brands" claim are unverified by independent reviews.

Mentions.so is an AI visibility and generative engine optimization (GEO) tracker that shows how often a brand appears in answers from ChatGPT, Perplexity, Claude, Gemini, Grok, DeepSeek, Llama and Google AI Overviews. It was built by the SEO agency Contact Studios, and that heritage shows. The product is shaped around the things an agency needs to put in front of a client: a clean dashboard, competitor comparisons, white-label reports and a pitch workspace.

The pitch is simple. You add prompts that matter to your business, pick competitors, and the platform runs them across AI models on a daily cycle. It then reports how often you are mentioned, where you land in the answer, how the model feels about you, and which domains the model cites. The homepage claims more than 100 brands use it. That number comes from the vendor, and I could not find independent reviews on G2 or Trustpilot to back it up.

Mentions.so is aimed at small teams, freelancers and agencies that want a cheap, readable view of AI search presence without paying enterprise prices. It also advertises a "Mentions 2.0" release in private beta, which promises more agent-driven functionality.

Key features

Visibility score and competitor ranking. The main dashboard shows a visibility score, defined as how frequently your brand appears in AI responses, with a trend against the previous period. A competitor table lists each brand with average position, visibility percentage, sentiment (a 0-100 style score) and share of voice. In the demo, the tracked brand sits at position 1 with 80.1% visibility, while rivals like HubSpot and Salesforce sit around 44.9%. It is a readable table and it answers the first question every client asks: how do we compare?

AI search share of voice and open web share of voice. Two separate views. AI share of voice shows your slice of brand mentions across model answers. Open web share of voice looks at the sources AI draws on (the demo says "27 sources analyzed"). Splitting these is useful because a brand can dominate the model's own knowledge and still be absent from the pages the model cites.

Live mentions feed. A running feed of real model answers, each tagged with the model, the prompt (for example, a question about a brand's mission in website visitor tracking) and the position where the brand was mentioned. Clicking through opens the prompt detail. This is handy for spot checks and for showing a client an actual answer instead of a number.

Mentions and citations over time. Daily counts of brand mentions and branded citations, with percentage change. Fine for trend reporting. The citation view is where I would want more depth: a competitor's review of the tool says citation position tracking is limited, and the site itself does not show per-citation rank the way more specialized tools do.

AI model performance. A per-model scorecard (ChatGPT, Meta, Claude and others) with visibility, sentiment and position for each. It makes it easy to say "we are strong in ChatGPT but weak in Gemini", which is the kind of statement clients respond to.

Topic overview and source domains.

• The topic matrix compares visibility score or position per topic across competitors, such as B2B data enrichment, lead scoring models and website visitor identification.
• Source domains lists the sites AI cites most often for your tracked queries, with the number of URLs, usage share, average citations and article types (how-to guide, landing page, ranked list, opinion).
• It flags whether a domain is you, a competitor or a non-competitor, and tags categories like social and forum. YouTube, LinkedIn and Reddit show up in the demo.

Insights Board. A Kanban-style board (Backburner, Ideas, To-do, Doing, Done) filled with prioritized recommendations. Examples from the demo: a high-value source worth outreach, a sitemap missing lastmod dates, and a sentiment narrative problem around pricing. It refreshes on a timer and items carry priority labels and types such as Source, Citation, Audit, Sentiment and Narrative. This is a decent idea. It turns a dashboard into a to-do list. But it stops at recommendations: you still write and publish the content yourself.

AI traffic and crawler analytics. The site promises traffic from AI models to your website and crawler analytics showing when bots run and where citation volume is growing. All paid plans include AI traffic analytics. The crawler section on the marketing page still carries placeholder copy, so I would test it in a trial before treating it as equivalent to proper server-log analysis.

Multi-language, multi-country tracking. Responses can be checked by country and language, with English, Portuguese, German, Dutch, Japanese, Hindi, Russian and Spanish shown. Dashboard filters cover model, competitor, location and topic.

Mentions Agent. A chat assistant over your data. You can ask it to chart this week against last week, explain how to improve sentiment, or list competitor features you lack. Useful for quick reports, though it is an analyst, not an executor.

Who is it for

The sweet spot is a small SEO or content agency managing 3 to 20 clients that wants to add an AI visibility line item to retainers. White-labeling, a custom domain and a pitch workspace on the $399 Agency plan are designed for exactly this. A founder or solo consultant tracking 25 to 50 prompts for one site also fits the $49 and $99 tiers, and unlimited seats mean nobody is charged per teammate.

It also suits SaaS marketing teams that just want to know where they stand against four or five competitors and do not need deep technical diagnostics.

Who should skip it? Larger in-house teams will likely find it agency-centric, as a competitor's comparison also notes. If you need to prove revenue impact through crawler logs, tie AI visits to Search Console and Bing data, scale content production, track shopping or ads placements, or monitor Reddit, YouTube and LinkedIn citations in depth, a fuller platform is a better match. Teams that want AI prompt suggestions will also be disappointed, since you bring your own prompts.

Integrations and ecosystem

Integration information is thin. I could not find a public integrations page, API documentation, or a changelog. The marketing page mentions connecting analytics, visibility and campaign data in the agent, but it does not name Google Search Console, GA4, Slack, Looker Studio or Zapier. Treat all of these as unconfirmed and ask before buying.

What is clear: there is a web app at app.mentions.so, white-label reporting and a custom domain on the Agency tier, and a pricing page that shows a monthly/yearly toggle. I found no browser extension, mobile app, or official GitHub, X or Discord presence. Agencies get free access to an "AI Search Community" on the top plan.

Pricing and value

The pricing section on the homepage lists four tiers. The yearly toggle takes about 16% off. A dedicated /pricing page returned a 404 during research, and the scraped text was partly garbled, so confirm the exact limits on the live page.

• Starter, $49/month: 25 prompts, daily updates, 1 site, unlimited seats, tracking for 3 LLMs, AI traffic analytics.
• Pro, $99/month (marked popular): 50 prompts, daily updates, 5 sites, unlimited seats, all LLMs, AI traffic analytics, tailored insights.
• Business, $199/month: 100 prompts, daily updates, 10 sites, unlimited seats, all LLMs, AI traffic analytics, tailored insights.
• Agency, $399/month (marked best value): everything in Business plus 300 prompts, a pitch workspace, unlimited sites, a custom domain, full white-labeling, priority support and community access.

I did not find a free plan or free trial, so check whether one exists. A competitor review reports that Google AI Mode is not covered on lower tiers.

How does that compare? Promptwatch's Essential plan is $95/month for 50 prompts, all AI models, 6,000 responses, a site audit, 360 Insights, content gap analysis, 500 Agent Credits, and MCP and API access, plus a free Explore tier and a 7-day trial on paid monthly plans. Its agency plans start at $199/month with unlimited projects and prompts. So Mentions.so is cheaper at the entry level, but you get less per dollar once you count crawler logs, content tools and optimization workflows. For a tracker-only need, the price is fair. For anything beyond tracking, the savings shrink.

Strengths and limitations

Strengths

• Low entry price of $49/month with unlimited seats on every plan.
• Clean, client-friendly dashboard with competitor ranking, sentiment and share of voice in one view.
• Strong agency features: white-label reports, custom domain, pitch workspace and unlimited sites on the top tier.
• Insights Board converts findings into a prioritized to-do list.
• Country and language filters for multi-market tracking.

Limitations

This is mostly a monitoring product, and the gaps against a full optimization platform are real. Promptwatch has these and Mentions.so, as far as I can tell, does not:

• Content generation and publishing. No Content Agent writing GEO content and pushing it to WordPress, Webflow or Framer, and no content refresh workflow.
• Content gap analysis. Recommendations exist, but nothing like a systematic answer gap report that turns missing prompts into briefs.
• IP-verified crawler logs. The crawler analytics section is vague, with no named bots (GPTBot, ClaudeBot, PerplexityBot) or crawl-to-citation path per page.
• Search Console and Bing data alongside AI data at the page level.
• Prompt intelligence. No prompt volumes, difficulty scores or query fan-outs, and no AI prompt suggestions.
• Reddit, YouTube and LinkedIn citation reports, AI Shopping, and Ads Radar.
• Real-UI coverage across 15+ engines, including Amazon Alexa and coding agents like Codex and Claude Code. Mentions.so lists eight model families, and the demo shows older model names such as Claude 3.5 Haiku and Grok 3 Mini Beta, which suggests the tracking may rely on API models rather than live interfaces.

Competitor LLM Pulse also says full AI response capture is missing, sentiment is basic, and benchmarking is shallow. That source is biased, but it matches what the marketing page shows.

Other worries: the testimonials are repeated and unverified, the site has leftover placeholder text, and I found no independent reviews.

Bottom line

Mentions.so is a sensible cheap tracker for agencies and small teams that want to start reporting on AI visibility without a big commitment. The Agency plan makes the most sense, because white-labeling and unlimited sites are where the product feels most finished. Verify integrations, model coverage and the trial situation before paying.

If you want to find out why you are not cited and then fix it, Promptwatch is the stronger choice. It adds crawler logs, traffic attribution, gap analysis, the Content Agent with CMS publishing and the Action Board on top of tracking, and it checks whether each fix worked. For other options, the GEO software directory at bestgeosoftware.com lists more trackers to compare.

Best use case: an SEO agency adding an affordable, white-label AI visibility report to client retainers.