How to measure your AI Share-of-Voice (honestly)

Everyone wants to know the same thing: which tools does ChatGPT recommend when someone asks it for one like mine? It's the right instinct and an unanswerable question. Nobody outside those companies can see what an assistant suggests inside a private conversation. Any tool that claims to measure "what ChatGPT recommends" is either guessing or quietly measuring something else.

So measure the something else — on purpose, and honestly. When a buyer researches the problem you solve, they run searches. The tools that show up in those results are the ones in the running. That's Share-of-Voice: not a peek behind the assistant's curtain, but a clean read on whether you're visible where people actually look. It's a proxy — and unlike the black box, it's one you can act on.

The method, in one breath

Pick the phrases your buyers type. Run them as web searches. Count which brands show up. Repeat weekly so you're watching a trend, not a screenshot.

That's the whole thing. The honesty lives in the details:

• Real queries, not vanity ones. "best tool to run an AI agent on a schedule," "Zapier alternative for AI agents," "automate a daily digest with AI" — buyer-intent language, not your product name.
• Count mentions in the results, not vibes. A brand counts when it's actually named in a result's title or snippet, once per result. Related-but-unnamed doesn't count.
• Track a fixed competitor set so week-over-week numbers are comparable.
• Keep the receipts. Every number should trace back to a result URL you can open. If it isn't checkable, it doesn't belong in the report.

What it can — and can't — tell you

It can tell you: where you're visible and where you're invisible, which queries you own versus which a competitor owns, and whether that's moving after you ship a comparison page or land a mention in a roundup. It's a scoreboard for your positioning in the exact place buyers do their research.

It can't tell you what any AI assistant recommends privately, or your true market share, or intent behind a search. Anyone who tells you their tool reveals ChatGPT's internal recommendations is selling you a number they made up. The measurable thing is visibility, and it's genuinely useful precisely because it's grounded.

What to do with the report

Read it as a map of openings:

1. Double down where you already show up. The queries you rank on are your real positioning — the language and problem-shapes where buyers connect you to the job. Lean into those in your pages and posts.
2. Treat a competitor's blind spots as gaps. A query where an incumbent is thin is an opening — a comparison, a how-to, a roundup inclusion. Ship it, then watch the next week's report to see the visibility move.
3. Trust the trend, not the snapshot. One run is noise. The week-over-week delta — a query you started appearing on, a competitor slipping — is the signal worth acting on.

Run it every week without lifting a finger

This is a perfect job for a scheduled agent, because the value is in the repetition and nobody wants to run twelve searches by hand every Monday. We packaged the method into a routine: AI Share-of-Voice weekly monitor. It runs the query set with built-in web search, tallies your brand against the competitors you name, and emails you a ranked table with checkable citations — and a plain note that it measures search visibility, not an assistant's private picks.

Copy it, swap in your brand and competitors and your buyers' real queries, pick a cadence, and let it run. Want the trend to compound? Attach a Google Sheet and have it append each week's table, or post the headline to Slack. Same agent, same data, now writing back to your tools.

That's Share-of-Voice done honestly: not a claim about what the machines think of you, but a weekly, checkable read on whether the people who need you can find you. Start a routine or browse the rest of the templates gallery.