Starts free.
Stays honest.
Free shows whether AI assistants name you. Starter and Pro check every day and hand you the files that fix the gaps.
| Prompt | ChatGPT | Gemini | Named instead |
|---|---|---|---|
| Best review software for contractors | cited | cited | — |
| Top review management providers | mentioned | missing | competitor-one.com |
| Review software for QuickBooks | missing | missing | competitor-two.com |
| Northlight vs alternatives | not measured | cited | — |
| Best review software for small teams | not measured | not measured | — |
Three plans.
Start one brand on Starter.
Monthly or yearly. Thirty days to change your mind, whole amount back.
Free
$0
no card
For finding out whether AI assistants know you exist.
- Three scans in any 30 days, no card
- Up to 5 questions, written from your homepage
- Checked on ChatGPT and Gemini
- See who AI names instead of you
- Citations checked on Gemini; ChatGPT not asked to look
- All six generators and three site checks
Creates your account. Three scans in any 30 days.
Starter
$49/mo
or $441 a year — three months free
$49$36.75/mo
$441 billed once a year — you keep $147
Citations measured on both assistants every plan covers, a daily check, and the fix files. Pro adds breadth. Start here, and move up any time.
For one brand that wants watching every day, and fixing.
- Switched on, not yet proven: Checked every day: Gemini daily, ChatGPT weeklynot yet proven
- Citations measured on ChatGPT and Gemini
- Fix files: schema, llms.txt, robots.txt, FAQ blocks
- Competitors named, with the questions they win
- History kept, so you watch it move
- A re-check you press, every seven days
Sign in, then pay on Paddle’s checkout from your account page.
Pro
$99/mo
or $891 a year
$99$74.25/mo
$891 billed once a year — you keep $297
For a brand competing across many questions and four assistants.
- Everything in Starter, across 40 questions
- Switched on, not yet proven: Adds Claude (monthly) and Perplexity (every third day)not yet proven
- Drafted pages for the questions you’re missing
- Switched on, not yet proven: An email when an AI answer about you changesnot yet proven
- Your whole team, no per-seat charge
Sign in, then pay on Paddle’s checkout from your account page.
Thirty days to change your mind. Write to us inside thirty days and the whole amount comes back — no conditions, and you do not have to say why. Cancel any time by one email or from your account page, and you keep access to the end of the period you paid for. Free needs no card.
Not yet proven, so tagged above: Starter’s schedule and the Pro alert email have never fired; Claude and Perplexity have not run twice yet. The daily run has fired every day since 6 Sep 2026 on our own Pro account. Everything else runs today, when you press the button.
You can do this by hand — we wrote the guide: a fixed question list, a spreadsheet and about two hours a month. Starter asks for you every day, keeps the history, and hands you the files.
Three steps, and no sales call.
- 01
Sign in or create an account
Free, no card. Takes an email address.
- 02
Pay on Paddle’s checkout
Paddle is merchant of record; we never see your card.
- 03
Run a scan, open the fixes
Your files are on the fixes screen; the daily check is scheduled from then on.
Example data — the shipping scan report on the invented brand Northlight: 8 questions across four assistants, each asked once — 32 checks, 30 came back, and the 2 that did not sit in the dashed tile. No customer’s scan appears anywhere on this site.
This is what a paid scan hands you.
On both paid plans every question is asked once per ask and the repeats come from asking every day — Gemini daily, ChatGPT weekly. Both assistants go and fetch — so cited is measured on both, where free measures it on Gemini only. Every tile carries its own denominator, and the dashed one reads 2 checks here — two runs that did not come back, sitting in that tile, outside every count, never a miss.
Prices are in US dollars.
Every plan, line by line.
Where something is not built or not yet proven, the cell says so.
| What it is | Free$0 | Starter$49/mo | Pro$99/mo |
|---|---|---|---|
| What it costs | $0 | $49 a month | $99 a month |
| Questions per scan | Up to 5, and the result says how many it asked | Up to 10. The extra questions over free are the other side of a cost cut we made on the free side, not what the money buys | Up to 40 — four times Starter’s |
| AI assistants covered | ChatGPT and Gemini | ChatGPT and Gemini | ChatGPT, Gemini, Claude and Perplexity — all four fetch, and the last two are only on this plan. They do not all answer the same way, and the difference is the thing worth reading. Most days Gemini answers from what it already holds — that tells you whether you are named. Once a week it goes and reads your site — that tells you whether you are used. Both are counted, separately, and never added together. ChatGPT answers weekly and fetches when it does, and Perplexity every third day — it cannot be asked without searching, so every one of its answers is a grounded one. Claude answers once a month, so on that assistant you get a presence check rather than a rate — and it is shown as one, never as a score out of one |
| Google AI Overviews | No | No | not yet proven The vendor key is on our server and the screen exists, and it has still never been run once — so it has produced no measurement at all. Treat it as not measurable until this cell says otherwise |
| Passes per question | One, and it is reported as one run | 1 per ask — and the repeats moved across days rather than inside one scan. Each question is asked every day, so it accumulates its runs over the month. Every figure carries the number of runs behind it | 1 per ask — and the repeats moved across days rather than inside one scan. Gemini answers every day, Perplexity every third day, ChatGPT once a week, Claude once a month, so a question accumulates its runs over the month. Every figure still carries the number of runs behind it |
| How often it runs | Only when you press the button. Nothing is scheduled on this plan and nothing is emailed | Press the button, plus a daily scheduled run — Gemini every day, ChatGPT once a week. not yet proven: the same daily scheduler has fired every day since 6 Sep 2026 on our own Pro account, and no Starter schedule has run yet | Press the button, plus a daily scheduled run — Gemini every day, Perplexity every third day, ChatGPT once a week, Claude once a month. The daily run has fired every day since 6 Sep 2026 on our own Pro account: Gemini each day, ChatGPT and Perplexity once each so far. not yet proven: Claude monthly and Perplexity every third day, until each has run twice on schedule |
| Run it again whenever you want | Three in any 30 days — the fourth is refused, naming Starter and the date the next frees up. The signed-out checker is a different surface with a different cap: 2 a day from one connection | Yes — one every seven days, up to 10 questions, not banked | Yes — one every seven days, up to 10 questions, not banked. The scheduled run above is what makes this plan continuous; this is the button for when you have just changed something |
| Times your questions get asked, per scan | 10 5 questions × 2 assistants × 1 pass |
20 10 × 2 assistants × 1 ask |
Not one number on this plan. Each question is asked once per ask, and how often it is asked depends on the assistant — Gemini daily, Perplexity every third day, ChatGPT weekly, Claude monthly. Your report shows the runs behind every figure rather than a single total |
| History you can look back through | No — you get the run you just did, with nothing kept to compare it against | Yes, kept | Yes, kept |
| Who is being recommended instead of you | Named, but not broken down | Named, and which questions they win | Named, and which questions they win — the same as Starter today. The split that runs is free against paid; there is no deeper Pro version of this behind it |
| The files that fix it — schema, llms.txt, robots.txt, FAQ blocks | No. Free tells you what is wrong | Yes — the same four files written from your own scan rather than from what you type into a generator, handed to you on the fixes screen | Yes, plus the drafts — the drafted page for a question you were missing from opens from the same fixes screen, assembled out of your own words with the gaps left blank. You can now type the answer into the blank yourself, on that screen: it is stored as your words, dated, and the draft is only marked ready once no blank is left. Nothing writes the answer for you, by design — there is no button that would |
| A report you can forward | Taking a copy asks for an email address. Nothing is sent to it but the one launch email, and only if you tick the box | Yes, downloadable | Yes, downloadable |
| A nudge when an AI answer about you changes | No | No | not yet proven — Pro only. When a new scan disagrees with the one before it, an email says what changed, which runs are compared, and what was measured. None has been sent yet |
| Connects to the other tools you use | No | No | Not built yet — and which tools is not decided either |
| Brands you can track | One | One | One |
| People from your team | As many as you like, no per-seat charge | As many as you like, no per-seat charge | As many as you like, no per-seat charge |
| The six generators, which run in your browser | All of them, no cap on runs | All of them, no cap on runs | All of them, no cap on runs |
| The three site checks, which fetch your site from our server | All three run | All three run | All three run |
- The six generators — FAQ, Organization, Product and Article schema, robots.txt and llms.txt — run in your browser and fetch nothing, so what you type into them is never sent anywhere. That is the only thing on this page that can be described that way.
- The three site checks — AI Crawler Checker, AI Readiness Score, Render Check — and the visibility scan all take a URL and fetch it from our server. They are server-side, they cost money per run, and they will carry a cap.
- The free caps — at most 5 questions, 2 assistants, one pass each — are our own working defaults and can tighten. Five is a ceiling, not a promise: the questions are written from your homepage, and the result tells you how many were asked.
- The signed-out checker at /ai-visibility-checker/ allows two scans a day from one connection — an office sharing one connection shares the two. A free account is three scans in any 30 days instead. There is also a daily ceiling on what we will pay the assistants; a refused scan is reported as not measured, never as a zero.
- Google’s AI Overviews is not measured on any plan today — the check has never run and there is no number, so nothing on this page is sold on it.
Example data — one item of seven from the shipping fixes screen on the invented brand Northlight. The screen is reachable from Starter up.
What a finding looks like once it reaches the fixes screen.
Each item states what to do and why, with the questions the finding derives from. From Starter up, the same screen hands you the files — schema, llms.txt, robots.txt and FAQ blocks — written from your own scan rather than from what you type into a generator.
We would rather leave a gap
than print a number we cannot keep.
Depth, not just more questions
A free scan asks up to five questions, Starter up to ten, Pro up to forty — but the count alone is not what the money is for. What the paid plans buy is repetition, and the two buy it differently: Starter asks each question once and comes back the next day; Pro does the same across four times the questions, so the repeats accumulate across the month instead of arriving three at a time. Either way the spread is reported, every figure carries the number of runs behind it, and the record is kept so you can compare.
Every cap in writing
The limits are on this page rather than in a terms document you meet at checkout. Where a check costs us money per run, it carries a number — and a scan a cap refuses is reported as not measured, never as a zero.
Your scan stays yours
Your brand, your competitors and where you are losing never appear in a public page, a case study or a shared dataset, and nothing is trained on them. If we ever want to use any of it, we ask you separately — never buried inside a terms page.
One rule sits under all three: a number we did not measure is a number we do not print. What that looks like in a result — the four states, the run counts, why an unrun check is never shown as a zero — is written down once, on how it works.
The awkward ones.
Including the ones that cost us a sale.
Free scan, as configured
Can you guarantee we’re named by ChatGPT?
No — and nobody can. No vendor controls what an AI model answers, so anyone promising a mention is guessing. What a paid plan buys is the measurement, recorded honestly, and the files that fix what those assistants read on your site.
Why pay when I could do this by hand?
You can — we wrote the guide: a fixed question list, a spreadsheet and about two hours a month. Starter asks Gemini every day and ChatGPT every week instead, keeps the history, measures whether each of them actually read your page, and hands you the schema, llms.txt, robots.txt and FAQ files to paste in.
Starter or Pro?
Starter, if you have one brand and care about ChatGPT and Gemini. Pro, if you need Claude and Perplexity too, or want more than ten questions watched. Competitor detail is the same on both. Switch from your account page and it takes effect at once; moving up is charged the difference for the rest of the period.
What if it isn’t worth it?
Write to us inside thirty days and the whole amount comes back. No conditions, you do not have to say why, and it does not matter how many scans you ran. One message, and a person reads it — no form and no retention flow. Your scans stay in your account, which drops to the free plan at the end of the period.
How do I cancel?
Write to hello@answermend.com and we will cancel it — that route always works. Your account page also opens Paddle’s own page, where you can cancel yourself, once there is a subscription for it to open. Cancelling ends the renewal, not the period: you keep access to the end of what you paid for, and nothing is charged after it.
Is anything not built yet?
Yes, and it is named. The scheduled runs have fired every day since 6 Sep 2026, on our own Pro account only; Starter’s schedule, Claude’s monthly run, Perplexity’s third-day run and the Pro alert email are not yet proven, and the cards say so. Google AI Overviews is built on Pro and has never been run, so nothing is sold on it. Integrations with other tools are not built.
Do I need a card to try it?
No. Free is three scans in any 30 days with no card, and it is not a trial that runs out. Paddle takes a card only when you choose Starter or Pro, and we never see it.
One email when something material changes.
Not a drip sequence. Both prices are on this page, the plans are on sale and the scan is running — so this list exists for the next thing that actually changes, not for a launch announcement.
Not sure yet? Check your site first.
Two scans a day from one connection, no account. If AI already names you everywhere, you may not need us.