Key Takeaways
- Read the audit as six parts, not one score. The board, the rivals, the pages being quoted, the page to write, the fix packs and thirty days of watching. Each part answers a different decision.
- Price it against your own hours, honestly. Asking 250 answers by hand at two minutes each is about eight hours before any analysis. If your hours are cheap and you only care about one AI tool, do it yourself.
- Notice the one part that is not in the box. Nobody at AnswerHalo publishes the page. The audit specifies it, and drafts it if you send five answers, but it goes live only when you put it live.
- Take the free check first. It asks ChatGPT the same kind of 25 buyer questions at $0. If it shows you are already named on most of them, the audit will mostly confirm it.
- Decide about the $300 a month at checkout, not later. Thirty days of watching come with the audit. Keeping it after that is $300 a month if added when you buy, or $399 a month once your baseline is no longer current.
Five hundred dollars is a real decision for a solo coach. It is a month of software, a weekend course, or a good chunk of a conference ticket.
So it is fair to ask what is actually in the box, part by part, and what the same work would cost in your own evenings if you did it by hand.
This is that inventory for the AnswerHalo audit, with the one part that is missing on purpose, and the situations where you should keep your $500.
What a $500 AI visibility audit buys you
A $500 AI visibility audit from AnswerHalo buys six things in one report link: the board, who gets named instead of you, the pages being quoted, the page for your worst question, fix packs, and thirty days of watching. It lands in your inbox by this time tomorrow.
What it does not buy is a published page or a promise of a mention. Those two limits matter as much as the six parts, and they are covered below rather than left to the small print.
For a general checklist to hold any vendor's audit against, see what a real AI visibility audit should include. This post is narrower. It is what our $500 actually buys and what each part is for.
The six parts, and the decision each one settles
Each part exists to settle one decision you would otherwise make on a guess: which questions you lose, who you lose them to, where to get mentioned, what to write first, what to fix off the page, and whether anything moved.
The board
25 questions, five AI tools, asked twice
Who is named instead
Only where two tools and both passes agree
The pages AI quotes
Ranked by how often
The page for your worst question
Every section in order, FAQ word for word
Action Packs
The off-site fixes your findings call for
Thirty days of watching
Re-asked every Monday, re-run at day 30
Publishing the page
The audit specifies it and can draft it. It goes live when you put it live.
The board. Your 25 buyer questions asked on ChatGPT, Perplexity, Google AI Overviews, Gemini and Claude. That is 125 checks, and each one is asked twice, so 250 answers are stored. The report shows the lower of the two passes, so one lucky mention never inflates your score. It settles which questions you lose, and whether you lose them on one tool or everywhere.
Who is named instead. The businesses that appear in your place. A rival is only named to you when two AI tools and both passes agree on it, because AnswerHalo's own test on August 1, 2026 asked an identical question set twice an hour apart and the rival lists did not overlap at all. This settles who you are really up against, rather than who happened to come up once.
The pages AI quotes. The sources the answers drew on, ranked by how often. Perplexity's help center describes answers that cite their sources, and Google's page on AI features in Search describes several related searches running behind one question. The list is where those searches keep landing. It settles where to get mentioned off your own site.
The page for your worst question. The address, the headline, every section in order, the FAQ block word for word, and which rival to concede rather than fight. Send five answers about your business, such as your published price and three dated outcomes with permission, and we draft the page for you inside three working days. It settles what to write first.
Action Packs. Guided instructions for the off-page fixes your findings call for, such as claiming profiles, making your details consistent, asking for reviews the right way, or a crawling problem on your own site. You receive the ones your results trigger, not all of them. This settles what to fix that is not a page.
Thirty days of watching. Your questions re-asked every Monday, then a full re-run at day 30 against the baseline taken on day one. It settles whether anything you changed actually moved.
Around all six sit a recorded walkthrough over your own numbers, a plan for the next thirty days built from that run's evidence, and a method section. Every figure in the report links to the stored answer it was counted from, with the question, the tool, the model, the time and the sources.
| Part | What it contains | The decision it settles | Illustrative hours by hand |
|---|---|---|---|
| The board | 25 questions on five AI tools, asked twice, 250 stored answers | Which questions you lose, and where | About 8 |
| Who is named instead | Rivals agreed by two tools and both passes | Who you are really up against | About 3 |
| The pages AI quotes | Sources ranked by how often they were used | Where to get mentioned off your site | About 2 |
| The page for your worst question | Address, headline, sections in order, FAQ word for word | What to write first | About 4 |
| Action Packs | Instructions for the off-page fixes your findings trigger | What to fix that is not a page | About 3 |
| Thirty days of watching | Weekly re-checks, then a full re-run at day 30 | Whether anything moved | About 6 |
| Publishing the page | Not included | When it goes live, which is yours | Yours |
What the same work costs in your own hours
Roughly 26 hours, on illustrative estimates. The only figure in that total that is arithmetic rather than judgment is the first: 250 answers at about two minutes each to ask, read and record is 500 minutes, a little over eight hours.
The rest are estimates, and yours could be faster or slower. Reconciling which names appear in both passes and on two tools takes a few hours with a spreadsheet. Tracing and ranking the sources takes a couple more. Writing a full page brief takes an afternoon if you have done one before, and much longer if you have not. Re-asking your questions every Monday for a month adds up quietly.
The part that is hardest to do by hand is not the typing. It is the discipline: asking the same frozen questions, in fresh sessions, twice, and not quietly rewording the one you lost. The Princeton-led GEO study found that how content is written changes how visible it is in generated answers, which cuts both ways: small wording changes in a question change the answer too, so a test you keep adjusting stops being a test.
If your time is worth more than about $20 an hour, the audit costs less than doing it yourself. If it is worth less, or you enjoy this kind of work, doing it by hand is a perfectly good choice, and the steps for a free version are in when a free AI visibility check is enough.
What $500 does not buy
It does not buy a published page, a guaranteed mention, or anyone doing the work on your site. The audit measures and specifies. Every change that could move an answer is one you make.
No published page. The brief is complete and the draft is available, but nothing goes live until you publish it. If you know you will not publish anything in the next month, the audit will sit in your inbox as an accurate description of a problem you did not fix.
No promised mention. Nobody can honestly promise that an AI tool will name you by a date, and we do not. What the report does give you is a way to check every number in it: each one links to the stored answer it was counted from.
No ongoing watching after day 30, unless you choose it. The watching included with the audit ends at thirty days. Keeping it is $300 a month if you add it while buying the audit, because a continuation runs against a baseline that is still current. Restarting later is $399 a month, because a restart needs a fresh baseline. The first charge is on day 31, since month one is already paid for.
No hands-on help at a published price. Done-with-you help with the work itself is quoted on request, which is why no number for it appears here.
When to skip it
Skip the audit if you have no real client results yet, if the free check already shows you named on most of your questions, or if nobody will publish anything in the next thirty days.
No results yet. The audit's page brief needs outcomes to put in it. A brand-new practice with no clients has nothing for an AI tool to repeat, and $500 will mostly confirm that. Come back once you have results worth publishing.
Already named. If the free check shows your name on most of the 25 questions, the audit will add detail rather than direction. And if the audit itself shows your mention count beating every competitor in the report, we tell you and refund the $500 before you ask.
Nobody will publish. This is the most common reason an audit is wasted money, and the damaging admission of this post. The audit is only worth its price in the hands of someone who will put one page live within a month. If that is not you right now, wait until it is.
Whichever way you lean, start with the free check. It asks ChatGPT 25 real buyer questions from your niche at $0 and emails a plain-English readout: your score, who got named in your place, and the one fix to start with. It runs as soon as you ask for it, and the report lands within the hour between 7am and 5pm Central, or by 10am the next morning outside those hours. For how the three options sit next to each other, see free check, audit and monthly watching compared.
Questions we hear the most
What does a $500 AI visibility audit buy?
At AnswerHalo, one report link covering 25 buyer questions across five AI tools asked twice, the businesses named in your place, the pages being quoted, the page for your worst question specified section by section, fix packs, and thirty days of weekly re-checks.
Which AI tools does the audit cover?
Five: ChatGPT, Perplexity, Google AI Overviews, Gemini and Claude. Each of your 25 questions is asked on all five, which is 125 checks, and every check is asked twice.
Why is every question asked twice?
Because the same question asked an hour apart can name different businesses. The report shows the lower of the two passes, and a rival is only named to you when two AI tools and both passes agree.
Does the audit write the page for me?
It specifies the page in full: the address, the headline, every section in order and the FAQ block word for word. Send five answers about your business and we draft it for you inside three working days. Publishing it is yours.
How long does the audit take to arrive?
It lands in your inbox by this time tomorrow, as one link to the report with a recorded walkthrough of your own numbers embedded in it.
Can I do the same work myself for free?
Much of it, yes. Asking and recording 250 answers takes roughly eight hours at two minutes each, before you reconcile names or trace sources. If one AI tool is all you care about, start with that.
What happens after the thirty days of watching?
Nothing, unless you choose to keep it. Continuing is $300 a month if you add it while buying the audit, because your baseline is current, or $399 a month to restart later. It is not a discount.
What if the audit shows I am already winning?
If your mention count beats every competitor in the report, we tell you and refund the $500 before you ask. Finding that out is worth knowing, and it is not worth charging for.
Your buyers are already asking. Find out what AI tells them.
The free check asks ChatGPT 25 real buyer questions about your niche and sends you the report within the hour, at $0. If it shows you are already getting named everywhere, we will say so in plain words.
Get my free check →