Case studies

Is Plausible visible in AI answers? We asked 11 buyer questions, it was missing from 4

SeenByLLMs · 2026-10-16 · An outside test of public AI answers. We are not affiliated with Plausible.

When buyers ask an AI model for privacy-friendly website analytics, does it name Plausible? We ran that test on 16 October 2026. The short answer: often, but not always. Plausible was absent from 4 of 11 answers, scored 61/100, and took 33% share of voice against 19% for the top rival, Matomo.

How we tested

We tested the category privacy-friendly website analytics with 11 questions written the way a buyer would type them into a chat box, not the way a marketer would phrase a keyword. They run from "Which website analytics should a small SaaS use instead of Google Analytics?" to "What analytics would you pick for a blog that cares about cookie banners?" We tracked five competitors: Google Analytics, Fathom, Umami, Matomo and Simple Analytics.

The model was Nemotron 3.5 Lightning (free) via OpenRouter, with OpenRouter web search switched on. This is not ChatGPT, Claude or Gemini, and none of those were tested here. We ran the 11 questions once, on 16 October 2026.

A brand counts as "named" when the answer contains the brand name or a known alias as plain text. Rank is the order in which a brand first appears in the answer. Pitch targets are pages the answers cited that name a rival but not the brand. We exclude competitor domains, the brand's own domain, and platforms such as GitHub, Reddit, YouTube and Wikipedia, because a pitch there is a different job.

Limits first: one model, one run, 11 questions. Read the numbers as a signal about what these answers look like on the day, not as a measurement of the market. One disclosure. For "Which analytics tools are best for EU visitors and privacy?" the model returned an empty answer. We count it as not mentioned, but it is a model glitch, not a competitor win.

Results, question by question

Every question was asked once. Here is each one, with Plausible's position in the answer.

The pattern is clean. Plausible wins where the question carries an intent the answer already has a shortlist for: GDPR alternatives, open source, static sites, cookieless, cookie banners. It ranked first in all of those, and in one the only tracked brand named was Plausible itself.

It loses where the question narrows to a stack, a budget or an audience the answer treats as niche. The Next.js question pulled in tools with Next.js install docs. The cheapest-tools and indie-maker questions pulled in small tools none of our five tracked rivals cover. Broad category questions are friendly to Plausible. Specific ones are not.

Where Plausible went missing

"What are the cheapest privacy-focused analytics tools?"

The answer named phpAnalytic, Loglesk, Muro Analytics, InsightFlare and Privatus Analytics. It named none of the five tracked competitors and did not name Plausible. This is a price-led question, and the answer reached for small tools whose whole pitch is cost, not for the better-known privacy players.

"Which privacy analytics tools work well with Next.js?"

The answer listed Simple Analytics at rank 1, then Beam, Sparklytics, Ours Privacy and Privatus Analytics. It cited Simple Analytics' own documentation page for the install. A framework question is answered with framework documentation, so the tools with an obvious Next.js install guide get pulled in and the rest never enter the list.

"Compare lightweight analytics tools for indie makers"

The answer named piratepx, MintyMetrics, Fusionaly, Hushstat and AnyAnalytics. Again, no tracked brand and no Plausible. "Compare" plus "indie makers" invites a list of small, cheap tools, and none of those five names appear anywhere else in the run.

"Which analytics tools are best for EU visitors and privacy?"

The free model returned an empty answer here. Nothing was named, including Plausible. The scorer counts that as absent, but no rival won it. Read the result as 3 real losses plus 1 empty answer.

The pages behind those answers

Three cited pages named a rival but not Plausible. All three came from the Next.js answer, and all three are other analytics vendors' own install guides:

• beam-privacy.com - Beam for Next.js - Privacy-first analytics for Next.js apps
• docs.oursprivacy.com - Integrate Ours Privacy with Next.js
• privatusanalytics.com - Next.js analytics: App and Pages Router

The page that put the one tracked rival in that answer was Simple Analytics' own guide, docs.simpleanalytics.com/install-simple-analytics-with-next. Nobody is going to list Plausible on a competitor's docs. The lesson is about format: for "works with framework X" questions, the model cites framework-specific install pages, so that is the page type Plausible needs to have and have found.

Why this matters

These are buying-intent questions. "Best X for Y" and "alternatives to Z" are what a person types when they are about to pick something, and the answer they get is a shortlist. If Plausible is not on that shortlist, the buyer never opens the site, never starts a trial, and analytics records nothing, because there was never a visit to lose. That is the quiet part of AI answers: the loss happens before the funnel, so it does not appear in any report. The three real losses here map to three question types: price (cheapest), stack (Next.js) and audience (indie makers). Each is a question a buyer asks right before choosing.

What Plausible could do next

1. Next.js gap. Make sure there is a dedicated, plainly titled "Plausible + Next.js" install page (App Router and Pages Router), the same format as the four vendor docs that won this answer, and link it from the main docs so search picks it up.
2. Price gap. Publish a page that answers "What are the cheapest privacy-focused analytics tools?" head-on, with real plan prices next to the alternatives. The winning answer was built from pricing pages (muroanalytics.com/pricing, phpanalytic.com, loglesk.io).
3. Indie-maker gap. Get onto comparison posts aimed at indie makers. The answer cited the tools' own sites (piratepx.com, mintymetrics.com, fusionaly.com, hushstat.com, anyanalytics.org), which suggests little editorial coverage exists for that exact query. A Plausible-written "lightweight analytics for indie makers" comparison would face thin competition.
4. Track more rivals. Add newer niche players such as Beam and Privatus Analytics to the competitor list for the next run. Thirteen untracked tool names showed up in the three real losses, so a 33% share of voice measured against the usual five likely flatters Plausible.

Run the same check on your brand

1. Create an OpenRouter key at openrouter.ai/keys, set a spend limit, and paste it into the app. The key stays in your browser.
2. Enter your brand, up to five competitors, and the questions your buyers ask, or start from the suggested prompts.
3. Press Run, then read which questions you lose, who was named instead, and which pages are worth pitching. Export HTML or CSV.

This run cost about $0.07 on our own OpenRouter key and took 0.7 minutes.

Run a free check