Answer-level visibility
See the exact questions, answers, and moments that triggered your ads.
About Metistic
A channel arrived without a way to check it.
Ads inside ChatGPT answers are bought like media and reported like nothing else in media. You get your spend and your impressions. You do not get to see which customer questions the money reached, or who was standing there when it didn't. Metistic exists to close that half.
- 100%Answer-level visibility
- Live marketsUS & UK. More coming.
- No gapsWe close the invisible half.
Real market coverage
Track ad delivery across the US and UK, with more markets coming soon.
Complete accountability
Close the gap between spend and influence with full transparency.
Built for this era
A new channel deserves a better measurement standard.
The problem
Every ad channel gets an independent check. This one doesn't yet.
Transparency shouldn't be optional. That's why we're building the independent layer conversational ads are missing.
SearchIndependent check exists
Search has had independent measurement for two decades. Nobody expects an advertiser to take a platform's word for their competitive position, because the platform reporting the numbers is also selling the inventory. The tooling that grew up around search wasn't hostile to it — it made the channel legible enough for large budgets to move into it with confidence.
Independent measurement · Verifiable reporting · Built-in trust
Conversational adsIndependent check missing
Conversational ads arrived without that layer. The reporting tells you what you spent and how often you were shown. It cannot tell you which questions you never appeared on, which advertiser holds the ones you lost, or what they promised the customer instead. Those are the questions that decide budget, and right now most advertisers have no way to answer them.
No independent layer · Incomplete visibility · Unanswered questions
Metistic exists to close that half. We bring independent truth to conversational ads.
Our principles
Four things we won't trade away.
01 Observation over estimation
Every number comes from a check that actually happened. Nothing on the platform is modelled, inferred from a proxy, or filled in where the data was thin. Where we haven't looked, we say so.
02 Absence is a finding
A question that carried no ad is a result, and it gets stored like one. Most measurement products throw those away, which is why they can tell you where you appeared and never where you didn't.
03 Sample size travels with the number
No rate appears anywhere — dashboard, report, export — without the count of checks behind it. A figure without its sample size invites a decision it can't support.
04 Geographic honesty
A market is measured from inside that market, on a real connection. Routing a request through a data centre and calling the result a country is the fastest way to produce numbers that look fine and mean nothing.
Integrity isn't a feature. It's our foundation.
Method
Ask the same questions, on a schedule, and write down everything.
- 01
A set per category
We build the questions customers actually ask on the way to buying in your category, and you edit them. That set is fixed, so week-to-week movement means something rather than reflecting a change in what we asked.
- 02
Checked daily, from the market
Each question is asked once a day from a real account on a real connection in the market being measured. Delivery varies between sessions, so a single check proves nothing and repetition is the whole method.
- 03
Every result recorded
Your ad, someone else's, or none at all — all three are stored with a timestamp and the creative captured. Rates are built from that record, and the record is always one click away.
It isn't complicated.
It's just done consistently, which is the part that takes work.
Our standards
Things we've decided against.
No modelled numbers.
If we didn't observe it, it isn't on your screen.
Coverage built on real markets.
We measure ad delivery using real accounts and real connections, so the results reflect what users actually see in each market.
No access to your ad account.
We measure from the outside. It's also why we can see everyone else.
No selling your data.
What we collect for you is yours. Aggregate benchmarks, if we ever publish them, will never be traceable to an advertiser's account.
No vanity metrics.
No scores, no grades, no index numbers built to look impressive and resist being checked.
No confident number too early.
Where the checks are too few, the interface says so instead of rounding uncertainty away.
Who we build for
Built for every decision that matters.
Brands advertising in ChatGPT
who want to know what the budget reached, not just what it cost.
Brands who aren't yet
who want to see what a competitor is doing there before deciding whether to follow.
Agencies
who have to answer the question when a client asks it.
See what your category looks like from the outside.
Give us your domain and your market. We'll build the prompt set for your category and start checking it.
Build your prompt setfor your domain and market.
Build my prompt setSetup in seconds
Get started right away.
No card required
Start free, upgrade anytime.
Ad insights, not opinions
Real data from real prompts.
Free to startNo ad account access needed