Key Takeaways
- Treat any ranking guarantee as disqualifying. Nobody controls what an assistant says next, and the only people who promise it are the ones betting you will not check.
- Ask what gets measured before you ask what gets done. A service that cannot tell you your starting number cannot prove it moved anything later, whatever it publishes as well.
- Volume promises are a red flag, not a bonus. Pages per week is a number the seller controls completely, which is exactly why it gets promised instead of results.
- Insist on seeing the raw answers. Stored, dated answers are the only evidence that survives an argument, and any honest provider already keeps them.
- Judge the refund terms, not the confidence. What happens if nothing moves is the part of the offer that tells you what the seller actually believes.
A new category attracts new promises. Some of the offers landing in coaches' inboxes this year sound like they were written by someone who has never watched an assistant answer the same question twice.
A few of them are outright impossible. Others are technically deliverable and still tell you nothing about whether your buyers will hear your name.
Here is the line between the two, in plain terms, including what we will not promise you.
What can an AI visibility service actually promise?
Three things: measurement, work, and honest reporting. It can tell you how often you are named today, do specific published work to change that, and show you the same measurement again later, whatever it says.
Everything past that line depends on systems nobody in this industry operates. Google puts it bluntly in its guidance on hiring a search professional: no one can guarantee a number one ranking. AI answers are generated fresh each time from sources fetched at the moment of asking, so they are less predictable than rankings, not more.
That is not a reason to do nothing. It is a reason to buy work and measurement rather than outcomes somebody else controls.
Five things nobody can promise you
A rank, a citation by a date, control of the wording, a fixed number of mentions, and any figure produced by a method they will not show you. If one of those appears in a proposal, the proposal is the finding.
- We will get you ranked number one in ChatGPT We will measure how often you are named today, and again in 90 days
- Guaranteed citations within 30 days We will publish the pages that make a citation possible, and show you which ones get read
- We control what the AI says about your business We change the evidence it reads, then we check what changed in the answers
- Ten optimized pages every week The pages your own test runs show are missing, in the order they matter
- Our proprietary algorithm gets you in the answers Here are the questions we ask, the tools we ask, and the raw answers we stored
The rewritten version of each promise is not the weaker one. It is the one you can hold somebody to, because every part of it is either work they perform or a number they can show you.
The fifth one is the quiet killer. A provider who reports a score without showing the questions behind it is asking you to trust a number they invented, and a number you cannot reproduce is not a measurement.
Why these are impossible, not merely risky
Because the answer is generated, not stored. The tool picks its words as it writes and pulls sources when you ask, which is why the same question twice in one afternoon can return different names, as covered in how to tell if your AI visibility is improving.
Two more reasons sit underneath. The evidence lives on other people's websites, so anyone promising a mention on a site they do not own is promising an editor's decision. And the research on nudging these systems, including the Princeton generative engine optimization study, reports shifts in visibility from better sourcing and clearer writing, not a lever that puts a chosen name in a chosen answer.
Google describes its own AI features as assembled from what its systems have indexed. The honest reading of that is simple: you can change what exists to be read, and then you find out.
| The promise | What it would require | Deliverable | Ask this instead |
|---|---|---|---|
| Number one in ChatGPT | Control of a generated answer | No | How often am I named today, out of how many asks |
| A citation by a fixed date | Another site's publishing calendar | No | Which sites will you pitch, and what have you placed before |
| We control what AI says about you | Access to a model's output | No | What evidence will you change, and where does it live |
| Ten pages a week | Only their own writing time | Yes, and it proves nothing | Which pages, chosen from which measurement |
| A visibility score of 90 | A method they will show you | Only if the method is published | Show me the questions and the stored answers |
| Results or your money back | A defined result and a date | Only when both are written down | What exactly triggers the refund, in writing |
What an honest commitment sounds like
It names a measurement, a piece of work, and a reporting date, and it says what happens when the numbers do not move. That is a smaller promise than the ones being sold, and it is the only kind anyone can keep.
Ours, since it is fair to ask: we will not promise you a citation by a date, a rank in any assistant, or a number that depends on somebody else hitting publish. What we do commit to is a baseline you can see, the work that changes what these tools can read about you, and the same questions asked again so the before and after sit side by side.
We also tell people when they do not need us. If your audit shows you already out-place every competitor, we say so and refund the $500 rather than selling you a plan. Whether a paid audit is worth it at all is argued both ways in is a paid AI visibility audit worth it, and the honest comparison with software you could run yourself is in monitoring tools versus done-for-you work.
What to do if someone promises it anyway
Ask four questions and read the answers rather than the confidence. What is my number today, how did you get it, what happens if it does not move, and can I see the raw answers from a previous engagement.
A provider who measures properly answers all four in a paragraph each. A provider who does not will change the subject to their process, their proprietary system, or how many clients they have. That switch is the tell, and it is worth more than any reference.
Then check the refund terms with the same eye. "Results guaranteed" with no defined result is a sentence, not a term. A named metric, a named date and a stated refund is a commitment, and it is perfectly reasonable to ask for it in writing before money moves. The four checks worth running on any provider, and the same four run on ourselves in public, are in who in AI SEO is legitimate, and how do you tell.
If you want your own number before you talk to anyone at all, the free check runs 25 real buyer questions from your niche through ChatGPT at $0 and emails a plain-English readout within the hour. It runs as soon as you ask for it.
Questions we hear the most
Can anyone guarantee my business will appear in ChatGPT answers?
No. Nobody controls what an assistant generates, the same way nobody controls search rankings. Google states plainly that no one can guarantee a number one ranking, and AI answers are less predictable than rankings, not more.
Is a guaranteed citation in 30 days possible?
Not as a promise. Getting mentioned on a site an assistant already reads is real work with a real chance of landing, but the date is outside anyone's control because publishing decisions belong to other people.
What about services that promise a set number of pages a month?
That is a promise about their own output, not your results. It is deliverable and it is easy to verify, so treat it as a description of the work rather than as evidence that the work moves anything.
How do I check a provider's claims before I pay?
Ask for the questions they will ask, the tools they will use, and a sample of stored raw answers from another engagement. Any provider who measures properly can show all three without hesitating.
Are money-back guarantees a good sign?
They are a good sign only when the trigger is written down. "Results or your money back" with no defined measure is theater. A named metric, a named date, and a stated refund is a commitment.
What does AnswerHalo refuse to promise?
A citation by a fixed date, a rank in any assistant, and any number that depends on somebody else publishing. We promise measurement, the work behind it, and honest reporting when the numbers do not move.
What does the free 25-question check include?
It asks ChatGPT 25 real buyer questions from your niche at $0, then emails a plain-English readout within the hour: how often you were named, who was named in your place, and the first thing to fix. It runs as soon as you ask for it.
What do I actually get for the $500 audit?
It runs 125 checks across five AI tools, names who is being recommended in your place, lists the sites those answers cite, and lands in your inbox by this time tomorrow. If it shows you already out-place every competitor, we say so and refund it.
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 →