Generative Pulse Goes Standalone: The 84% Stat Behind AI Visibility

The short version: Muck Rack has spun its AI visibility tool out of the PR platform. Generative Pulse opened early access today, October 26, 2026, as a standalone product for brand, marketing, PR, content, SEO and executive teams — and you no longer need a Muck Rack subscription to try it. The pitch is simple: it is no longer enough to know whether your brand appears in an AI answer. What matters is what the answer says about you, whether it gets the facts right, and which sources shaped it.

That is a real shift in how these tools are sold. Until now, Generative Pulse lived inside Muck Rack, built for PR teams. The standalone version is aimed at every team that owns a piece of the answer — comms, brand, marketing, content, SEO and the executive floor. The original PR-focused tool is not going away: it is now called Muck Rack AI Visibility, and it lives in a new AI Visibility tab in Muck Rack's Coverage Hub. Muck Rack says it will keep investing in both, and that plenty of customers will use both.

What Generative Pulse actually does

Generative Pulse measures how AI portrays a brand across ChatGPT, Gemini, Claude, Perplexity and Google AI Overviews, and it does it across locations and audiences — because, as Muck Rack puts it, AI does not give everyone the same answer.

The measurement runs on a framework the company calls SEEN, four signals that are worth spelling out because they are the whole product:

• Share of answer — whether AI includes your brand at all when the question comes up.
• Evaluation — how favourably AI assesses your brand against competitors.
• Error rate — whether the claims AI makes about you match your facts.
• Narrative alignment — whether AI carries the messages you actually want out in the world.

Showing up is the starting line; the other three signals are where the product earns its keep. Each answer is traced back to the claims, citations and sources behind it, including the journalists and outlets in Muck Rack's media database. The findings then turn into prioritised recommendations across earned media, owned content, technical access and monitoring — each with an owner and a measurable outcome attached. As stories break and people start asking AI new questions, the platform suggests what to monitor next, using Muck Rack media data and what the company describes as millions of real AI conversations each month. Teams decide which suggestions to accept.

The public homepage shows the idea in one screenshot: a demo dashboard for a fictional airline, with share of answers at 86%, evaluation at 74%, error rate at 8% of claims needing correction, and narrative adoption at 42% — each plotted against two named competitors. Those are demo figures, not a real brand's results, but they show what a buyer actually gets to look at: four numbers, a competitive set, and a trend line.

Why Muck Rack spun it out

The company history matters here, because it explains why a PR platform is the one selling this. Muck Rack was started in 2009 by Greg Galant and Lee Semel as a place for journalists to connect; more than 10,000 journalists asked to be added in the first year. PR professionals started using it to find the right journalists, so a platform for communications teams followed in 2011. The company ran without outside funding for 13 years, then took its first institutional investment in 2022 — a $180 million Series A from Susquehanna Growth Equity — and says it has stayed profitable and founder-controlled throughout.

Generative Pulse launched in 2025 inside Muck Rack, alongside the company's "What Is AI Reading?" research. The lesson from customers, according to Galant's announcement post, is that how AI describes a brand is not one team's problem: comms, brand, marketing, content, SEO and executive teams all own a piece of the answer. A tool built only for PR teams could not serve the rest of them, so the product gets its own brand and its own platform, built around owned content, earned media and technical work in one view, on top of the same journalist and coverage data behind Muck Rack.

"Earned media is the main source AI draws on when it describes a company," Galant said in the launch announcement. "Teams need to know more than just that they showed up in an AI answer. They need to know what AI said, the evidence behind it, and whether owned, earned or technical work will change it."

The 84% number that makes the case

The whole category rests on one striking finding, and it is worth checking before you buy into it. Muck Rack's May 2026 edition of "What Is AI Reading?" analysed more than 25 million links cited by ChatGPT, Claude and Gemini across 17 industries, and found that earned media — coverage a brand neither owns nor pays for — accounts for 84% of the sources AI cites. Paid and advertorial content together barely register, at 0.3%. Journalism alone accounts for around 27% of citations, consistently across every edition of the study.

Two caveats belong in the same paragraph. First, Muck Rack's definition of earned media is broad: it counts journalism, academic research, government sources, encyclopedic sites and third-party corporate content, so 84% does not mean 84% of citations come from news stories. Second, this is vendor research — Muck Rack sells the product behind the data. The pattern holds across three editions (the July 2025 and December 2025 editions showed 95% and 94% non-paid citations), and independent work points the same way: a September 2025 arXiv preprint found AI search systems show "a systematic and overwhelming bias towards earned media" over brand-owned and social content, and McKinsey has put a brand's own sites at just 5 to 10 percent of the sources AI search references. Ahrefs, studying 75,000 brands, found branded web mentions correlate with AI visibility at 0.66 to 0.71 — a correlation, not proof of cause, but it supports the same working view.

One more finding from the December 2025 edition is the one that should worry PR teams most: only a 2% overlap between the journalists brands pitch most and the journalists AI systems cite most. Pitching the wrong list is not just wasted effort any more; it is invisible effort. Press-release citations rose fivefold between July and December 2025, with releases carrying more statistics, bullet points, action verbs and objective sentences cited at higher rates — a detail that reads like a how-to for anyone writing announcements now.

The market context is moving the same way. Forrester's 2026 research on business buying found generative AI is now the starting point for B2B buyers, who rate AI and conversational search as a more meaningful source than a vendor's own website or sales team. The shift is not limited to search: personal AI agents are becoming the first place people ask questions, from Meta's Muse to the AI agents now living in text messages. Muck Rack's own State of PR 2026 survey found nearly three-quarters of PR professionals say generative engine optimisation (GEO) is at least somewhat important to their communications strategy — yet 29% say no one at their organisation actually owns it. That ownership gap is exactly the buyer Generative Pulse is aimed at.

What it costs — and what it doesn't say yet

Here is the honest gap in today's announcement: Generative Pulse has published no price. Early access is request-only at generativepulse.ai, with no public tiers, no per-prompt rates and no free plan described anywhere on the site. The full launch is set for early 2027, and the company will show the product live at the Generative Pulse Summit on October 28 in New York City, alongside new "What Is AI Reading?" research. Until pricing appears, the only honest comparison is against what rivals publish — and that comparison is instructive, because this category's pricing is a maze of prompt caps, engine add-ons and quote-only enterprise tiers.

Before the table, three pricing fine-print points that apply to the whole category. First, a "prompt" is not a standard unit: it can mean one tracked question refreshed daily, or one query run, and that difference changes what you get by a factor of several. Second, the headline price rarely includes every engine your buyers use — several vendors charge extra per engine. Third, annual billing usually carries a discount (Otterly, for example, lists 15% off annual billing), so monthly figures are the ceiling, not the floor.

Generative Pulse vs the field

Two things stand out. First, the spread is enormous: $29 a month gets you 15 prompts on four engines at Otterly, while Profound no longer publishes any brand price at all — its pricing page lists only a free 7-day trial and a custom Enterprise plan behind a demo, with agencies paying $99 a month plus $399 per client workspace. Third-party trackers put its enterprise deployments at $2,000 to $5,000+ a month. Second, Generative Pulse joins Profound in publishing no brand price at all — the difference is that Profound at least lists an agency rate ($99 a month) and a free trial, while Generative Pulse lists nothing yet. That is normal for an early-access launch, but it means you cannot yet judge whether its journalist-and-coverage data is worth a premium over the self-serve field. Ask for the price, the prompt cap and the engine list in the same conversation — and get the annual-billing figure, not just the monthly one. Peec, for example, drops from $95 to $80 a month on annual billing, a 15% saving its own FAQ confirms.

Profound deserves a note of its own, because it is the category's reference point: it has raised $155 million in total funding, including a $96 million Series C at a $1 billion valuation led by Lightspeed, and is described in industry coverage as the first unicorn in the GEO category. It serves more than 10% of the Fortune 500, with named clients including Target, Walmart, Figma, MongoDB, Ramp and U.S. Bank, and was named a representative vendor in Gartner's first Market Guide for Answer Engine Visibility Tools. Its Conversation Explorer draws on panel data from a reported 1.3 billion user conversations. Generative Pulse's answer to that scale is different: not conversation volume, but the journalist and outlet data behind Muck Rack — the bet is that knowing who wrote the source matters more than knowing how many people asked.

Pros and cons

What works in its favour: the earned-media data is a genuine differentiator — no rival in the table can trace an AI answer back to the journalist who wrote the source. The four SEEN signals measure portrayal, not just presence, which is the question a brand team actually needs answered. The audience is right: GEO has no owner in 29% of organisations, and a tool built for brand, marketing and executive teams (not just PR) meets that gap. And the underlying research is unusually transparent for vendor work, with three published editions and a methodology you can argue with.

What to watch: there is no published price, no public trial terms and no independent hands-on review of the standalone product yet — everything above comes from the vendor's own announcement and site. The SEEN scores are vendor-defined; there is no public benchmark showing how accurately they predict real buyer behaviour. The 84% headline rests on a broad definition of earned media and on vendor research. And the product is early access only, with a full launch months away — expect rough edges and changing terms.

Who should request access — and what to ask

Request early access if you are the person who gets asked "what does ChatGPT say about us?" and you currently have no answer beyond a screenshot. That is most often a brand, content or SEO lead at a company with real earned-media coverage to work with — the product's value collapses if there is nothing for AI to cite. If your brand has almost no third-party coverage, the honest first step is earning some, not buying a dashboard to watch the gap.

When you get the demo, ask five things: the price and what a prompt cap buys; which engines and regions are included at that price; whether the SEEN scores update daily; how the recommendation engine picks its actions and whether you can see the underlying answers; and what happens to your data if you leave. Then run the same ten questions through a $29 tool and through Generative Pulse for a month, and see which one changes a decision. A visibility score that never changes a decision is a subscription, not a tool.

FAQ

What is Generative Pulse?
A standalone AI visibility and portrayal platform from Muck Rack, launched into early access on October 26, 2026. It measures how AI systems describe and recommend a brand, traces answers to their sources, and recommends actions across earned media, owned content and technical work.

What is the SEEN framework?
Generative Pulse's four-signal scorecard: Share of answer (whether AI includes your brand), Evaluation (how favourably it rates you against competitors), Error rate (whether its claims match your facts) and Narrative alignment (whether it carries your messages). It runs across ChatGPT, Gemini, Claude, Perplexity and Google AI Overviews.

How much does Generative Pulse cost?
No price has been published. Early access is request-only at generativepulse.ai, ahead of a full launch in early 2027. Rival AI visibility tools that publish prices list entry tiers from $29 a month (Otterly.AI Lite) to $250 a month (Scrunch Core), with enterprise tiers quoted individually.

What is GEO, and is it different from SEO?
Generative engine optimisation (GEO) is the work of shaping what AI answers say about a brand; answer engine optimisation (AEO) is the closely related term. SEO makes pages findable in search results; GEO makes brands describable — accurately and favourably — inside AI-generated answers, where the sources are overwhelmingly third-party coverage rather than your own site.

Do I still need Muck Rack to use it?
No. The standalone platform does not require a Muck Rack subscription. Existing Muck Rack customers keep their AI visibility data inside Muck Rack, in a new AI Visibility tab in Coverage Hub, under the name Muck Rack AI Visibility.

Key takeaways

• Generative Pulse opened early access on October 26, 2026, as a standalone platform — no Muck Rack subscription required, full launch in early 2027.
• Its SEEN framework measures four things: whether AI includes your brand, how favourably it rates you, whether its facts are right, and whether it carries your message.
• The case for the category is Muck Rack's "What Is AI Reading?" research: earned media supplies 84% of the sources AI cites, while paid content supplies 0.3% — a vendor study, but one whose pattern independent research supports.
• No price is published yet, so judge it against rivals that do publish: entry tiers run from $29 a month at Otterly.AI to $250 at Scrunch, with Profound's enterprise pricing quoted per deal.
• The product is worth a look if your brand already has third-party coverage for AI to cite; if it has almost none, earning coverage comes before buying a dashboard to measure it.

Published October 26, 2026. Generative Pulse is in early access; pricing and terms may change before the full launch in early 2027. Prices for rival tools are as published or reported in October 2026 and should be confirmed on each vendor's pricing page before you buy.

Sources: Muck Rack, "What's next for Generative Pulse" (official announcement, October 26, 2026); GlobeNewswire launch release (October 26, 2026); generativepulse.ai (official product site); Profound pricing page (official — no brand price published, trial free, Enterprise custom; Agency Growth $99/mo + $399 per client workspace); Peec AI pricing page (official — Starter $95/mo, Pro $245, Advanced $495; annual from $80/mo); Scrunch pricing page (official — Core $250/mo); Search Influence, enterprise AI visibility platforms compared (Profound funding and Gartner guide); Contentstack, free AI visibility tools compared (Otterly and Scrunch pricing); AI Hunter, best AI visibility tracking tools 2026 (Otterly and Peec pricing); CMSWire, earned media as a data play (Forrester 2026, "What Is AI Reading?" findings).