How To Audit Whether AI Recommends Your Brand
More actions
Finally, pay attention to how they talk about their existing clients. Somebody who describes a client's category accurately, names the specific constraint that made the work difficult, and mentions something that did not work has actually done the job. Somebody who describes every engagement as a success in identical language has either been unusually lucky or is describing a template.
There is also a mechanical problem. Manufactured mentions tend to be uniform in language and timing, which is exactly the pattern that gets discounted. The effort produces a body of sources that agree suspiciously well and carry less weight than a smaller number of genuine ones.
Set Up So You Do Not Fool Yourself Open a signed out session, or a fresh one with memory and personalisation disabled. This matters more than anything else in the method. An account that has spent the week researching your own company will show you a flattering picture that has nothing to do with what a stranger sees.
Present but described wrongly means a source problem, and the source list tells you which page to correct. Present and accurate on definitional prompts but absent on the who should I hire prompts means your category presence is fine and your commercial positioning is not corroborated anywhere independent.
Two asking who to hire or buy from for the thing you sell. Two describing the problem your product solves without naming the category. Two comparing named competitors. Two asking about a specific situation your best customers are in. One asking directly who your company is. One asking whether your company is any good.
You can do this yourself in about half an hour, with no subscriptions and no technical knowledge. It will not be as thorough as a full engagement, and it is more than enough to establish whether you have a problem and roughly what kind.
This is usually a few days of work, it frequently explains a poor baseline entirely, and it improves conventional search as a side effect. It is the cheapest part of the whole exercise and the most commonly skipped.
What Not to Do, and Why It Backfires Fabricated reviews, seeded forum threads under false identities, and paid placements presented as independent all exist and all fail on the same axis. Detection has improved, platforms enforce against it, and the reputational cost when it surfaces exceeds anything the visibility was worth.
What Not to Do in the Name of Legibility Hidden text intended only for machines fails on every axis. It is detectable, it violates most guidelines, and it produces exactly the uniform low quality signal you were trying to avoid.
What You Can Do Legitimately More than most teams assume. Claim every profile that allows it and complete it properly. Correct factual errors on platforms that accept corrections, which most do when you have evidence. Respond to reviews, including critical ones, since an unanswered complaint reads as inattention.
Good versions read like this: mention rate on evaluation prompts rose from two in fifteen to six in fifteen, which we attribute to the three directory corrections completed in week two, though a competitor also stopped publishing during the same period.
Read the Source List Before the Prose Where citations are shown, list every domain and count how to get recommended by AI assistants often each appears. This is the single most useful output of the whole exercise, and most people skip it because the prose is more interesting.
The practical result is that a claim appearing only on your website is treated as a claim, while the same claim appearing in a trade publication, a review platform and a forum thread starts being treated as a fact about the world.
Refusals matter. A report that only contains successes is either describing a suspiciously easy month or omitting the parts that did not work, and the omitted parts are usually where the useful information is.
Ask to See a Prompt Set The first question is the most revealing. Ask them to show you the prompt set from a current or recent client, with the client's name removed. A team doing real work has this and will show it, because the prompts are craft rather than secret sauce.
Ask What They Cannot Measure A competent practitioner will volunteer limitations before you ask. Assistant answers vary between sessions. Referral attribution is inconsistent. Some assistants cannot be measured reliably at all. Sample sizes in the published research are small.
This section sounds procedural and it is the foundation of everything after it. A prompt set quietly edited between runs makes every trend line in the document meaningless, and it is the easiest way to manufacture improvement without doing anything.
The Rendering Question This is the one real technical constraint. Content that only exists after JavaScript executes may be invisible to a retrieval fetch, which is not a browsing session and does not always run scripts.
Vague answers about digital PR are a warning sign. Good answers are concrete: they have read your baseline source list, they know which platforms allow corrections, they have a view on which comparison articles are worth approaching, and they will tell you which ones are out of reach.