The AI Search Visibility Briefing28 October 2026

Semrush removes the listed price from its top AI visibility tier

The Advanced tier of Semrush One now reads 'Custom' and 'Request demo', with unlimited daily prompt tracking. It is the clearest sign yet that AI visibility is being sold as an enterprise line item, not a metered add-on.

Unlimited prompts to track daily

— Semrush AI Toolkit
Goodie AIRankscaleMentionableEvertuneAhrefs Brand Radar

This Week's Headline Move

The most consequential AI visibility vendor move of the fortnight is a pricing page change. Semrush One's Advanced tier no longer lists a self-serve annual price, previously $549 per month, and now reads 'Custom' with a 'Request demo' button. In the same edit, its AI Visibility prompt tracking changed to 'Unlimited prompts to track daily'. The tier also adds Team users, Reporting and Advanced API data integration.

Read together, those three changes describe a repositioning. Prompt volume stops being the meter. Seats, reporting and API access become the reasons to buy the top tier. That is an enterprise software motion, not a rank-tracker motion. It also puts pressure on every vendor still selling AI visibility by the prompt count.

Semrush paired the pricing change with the H2 edition of its AI Visibility Index, built on 180M+ AI answers across 22 industries and adding a first UK benchmark. The study and the pricing page tell the same story: Semrush wants to own the benchmark narrative and the enterprise budget line that follows it. For buyers, the practical question is what 'unlimited' covers: engines, refresh frequency, seats, and historical depth.

Three More Signals Worth Watching

  • Evertune opened a self-serve entry point at $60/month. The new Starter plan sits below the sales-gated $800/month Pro tier, takes a credit card, and includes a ChatGPT Ads agent. It is a deliberate wedge for brands and agencies that will not book a demo to test AI visibility.
  • Mentionable published a three-tier structure at €79, €149 and €299. All tiers include the six tracked AI engines and pooled scan credits. MCP, REST API and Looker Studio sit on Pro and Agency, while white-label reporting is reserved for Agency. The split separates measurement from agency-grade reporting.
  • Rankscale widened its plan comparison beyond rank tracking. It now details ChatGPT follow-up forms, local and map-pack results, and AI-crawler visit analytics across tiers. Those are workflow features, not vanity metrics, and they signal where buyers are asking for depth.

The Pattern

The fortnight shows a market splitting along two axes at once. On price, vendors are moving in opposite directions: Evertune at $60/month and Mentionable from €79 are building low-friction, self-serve entry points, while Semrush pushes its top tier behind a demo gate at 'Custom'. On capability, the differentiator is shifting from prompt counts to what happens after the dashboard.

Ahrefs Brand Radar made the same shift in its own copy, replacing 'Make AI recommend your brand' and 'Turn SEO into AEO' with 'Study how your brand shows up in AI' and a revenue-channel pitch. It also dropped YouTube and Reddit from the hero subhead. Scrunch published two articles, one drawing on a 600+ respondent survey, framing action over dashboards. Goodie AI went further, publishing a comparison page that frames AthenaHQ as 'enterprise-gated GEO' against its own 'full-stack enterprise AEO'.

Taken together, these moves say that tracking is becoming table stakes. The fight is now over follow-up prompts, local results, crawler visits, API and MCP access, white-label reporting, and attribution. Vendors are repositioning around the work that follows the chart, because that is where budgets survive scrutiny.

What SEO, Content, Brand and PMM Teams Should Do This Fortnight

  1. Audit your prompt caps before renewal. If your plan has a fixed prompt count, model what happens when you add Claude, Gemini, Perplexity and follow-up prompts. Ask vendors whether 'unlimited' means all engines, daily refresh, and named seats, or only a single engine at a low refresh rate.
  2. Check reporting and API access on your current tier. MCP, REST, Looker Studio and white-label reporting are increasingly gated. If your team delivers client or executive reporting, confirm the tier includes those outputs before you sign.
  3. Add follow-up and local intent to your test plan. Rankscale's update is a reminder that answer engines handle multi-turn and local queries differently. Run a small set of follow-up prompts and compare citations, not just initial rankings.

Closing CTA

If you want a quick read on where your own brand stands before your next vendor conversation, start there. IndustryLens gives you a free starting point so you can compare tools with your own data in hand. Check your own brand's AI search visibility, free. Run the Free CI Pricing Agent and see where you stand: https://industry-lens.com/free

Track your own competitors — not just ours

IndustryLens turns competitor activity across 350+ sources into a briefing like this one, for your competitive set — every week, no manual research.

Start free trial

← Back to the weekly briefing archive · Explore the AI Search Visibility Briefing hub