the short version
Do it yourself once to find out whether you have a problem. Buy something when you need the answer to be repeatable, comparable and forwardable.
For a one-off gut check, do this. It costs nothing and it will tell you whether the problem is real. Come back when you need it to be repeatable.
Complete, for the first question. Opening ChatGPT and typing your category in tells you whether the problem is real, which is the only thing that matters at the start, and it costs nothing. We would rather you did that than bought anything. Everything on this page is about the second question and the tenth, not the first.
| dimension | unbound geo | Asking ChatGPT yourself |
|---|---|---|
| Cost | unbound geoFree for the first scan, then from 19 USD a month. | Asking ChatGPT yourselfFree. |
| Repeatablechanges the decision | unbound geoThe same prompt set, the same market, the same weights, so two runs are comparable. | Asking ChatGPT yourselfNo. Answers move, and you cannot tell a real change from normal variation. |
| Personalisationchanges the decision | unbound geoClean calls with no account history, geolocated to the market you choose. | Asking ChatGPT yourselfYour session, your history, your location. You are not seeing what a buyer sees. |
| Evidence for someone elsechanges the decision | unbound geoA dated report at a permanent URL that you can forward. | Asking ChatGPT yourselfA screenshot. |
| Sources | unbound geoEvery cited URL, ranked by how often the engines read it. That is the press target list. | Asking ChatGPT yourselfYou can read the citations one at a time. |
No scoring matrix and no weighted rubric. Most of these decisions turn on one fact about how you work, so the honest format is a list of facts and what each one implies.
pick Asking ChatGPT yourself
pick unbound geo
Every line above points at exactly one option, including the ones that point away from us. A decision guide where every branch ends at the same product is not a guide, it is a funnel with headings.
You can, and for a first check you should. What you cannot do is trust the answer as a measurement. Your session carries your account history and your location, so the answer is friendlier to you than the one a stranger in your target market gets, and asking twice returns two different results with no way to tell a real change from normal variation.
Different enough to matter. Personalisation, chat history and geolocation all shape the response. unbound geo makes clean calls with no account history, geolocated to the market you choose, using the same prompt set and the same weights each time so two runs are actually comparable.
Yes, plus your time. Our free scan is also free and needs no card; it exists precisely because the manual check is the reasonable first step and we would rather meet you there than argue with it.
A dated report at a permanent URL that says the same thing in two weeks, which is what lets it travel inside a company with nobody there to explain it. Plus every cited URL the engines read, ranked by how often they read it, which is a press and content target list rather than a set of citations you open one at a time.
A single scan covers five prompts on the free tier and up to thirty on the top tier, across up to four engines, with every raw answer stored. Matching that by hand means running the same questions in the same order in a clean session per engine, recording each answer, and repeating it on a schedule. Most teams do it once and never again, which is the actual problem.
Pricing noted for Asking ChatGPT yourself: Free, plus your time. Pricing changes and we do not monitor theirs, so check their site.
Written 30 July 2026. If anything here is wrong, write to hello@withunbound.com and we will correct it.
Five real questions against a live engine with web search on. You get the answer text, who got named, and the sources the engine read. No account, and it takes about a minute.