Nobody Reads Your Résumé Anymore. They Ask an AI About You.
A referral used to end with a document. Someone asked who you were, and you sent something — a résumé, a capabilities deck, a link to your site. You controlled the artifact. You chose what it said.
That step has quietly disappeared. The person checking you out now opens ChatGPT, or Gemini, or Perplexity, and types a question. What comes back is not your document. It is a paragraph assembled from whatever the web happens to say about you, written by a system you have never spoken to, delivered to someone who will never see the sources.
And if that system has nothing to say about you, it does not say so. It simply answers with someone else's name.
There is no page two of an AI answer
This is the part that most business owners have not absorbed yet, and it is the whole game.
Google gave you a long tail of survival. Ten results on page one, ten more on page two, and a searcher who scrolled, opened four tabs, and compared. Ranking eighth was not good, but eighth was visible. Eighth got clicked. Plenty of businesses were built on eighth.
An AI assistant does not work that way. Ask it for a dentist in your town and it names three. Maybe four. There is no eighth position, because there is no list long enough to have one. The long tail did not get shorter — it got deleted.
So the question stopped being where do I rank. The question is now binary. Either the model reaches for your name, or it does not, and the gap between those two states is the whole of your visibility.
The omission is silent, and that is what makes it dangerous
When your Google ranking slipped, you could see it. You opened Search Console, watched the line go down, and knew you had a problem.
Nothing tells you that an assistant left you out. The customer does not know they were shown a partial list — the answer arrives with the confidence of a complete one. You do not know either. There is no notification, no drop in a chart, no rejection email. The phone simply rings less, and it is very easy to attribute that to the economy, or the season, or the market.
Businesses are losing this quietly, right now, and most of them have not connected the two facts.
Five systems, not one
It is tempting to think of this as “the ChatGPT problem.” It is not. Five systems now stand between most businesses and most new customers: ChatGPT, Google's AI Overviews and Gemini, Perplexity, Claude, and Microsoft Copilot.
They do not share an index and they do not agree. I have run the same question across all five for the same business and had it named confidently by two, mentioned in passing by one, and omitted entirely by the other two. Being known to one is not being known. Each has its own view of who exists, and each is somebody's default.
Why an assistant names one business and not another
The useful thing about these systems is that they are not mysterious, and they are not for sale.
There is no ad slot in an AI recommendation. You cannot outbid your way into the answer. What decides it is whether the model can find the same facts about you, stated consistently, across several sources that do not belong to you.
A model builds its answer from corroboration. Your own website is a single claim — necessary, but it is you talking about yourself. What gives that claim weight is everything that agrees with it: the directory listing with the same address, the review platform with the same phone number, the trade publication that quoted you, the association page that lists you, the structured data that states plainly what you do and where.
When those agree, the model has a fact. When they conflict — an old suite number here, a disconnected phone there, a business name that shortened three years ago but not everywhere — the model has ambiguity, and it resolves ambiguity by moving on to a competitor it is more sure about.
That is the good news buried in all of this. Being recommended is not a budget. It is an engineering problem, and engineering problems are solvable by people who care enough to be precise.
This is not only a small-business problem
The résumé is the sharpest version of it, because professionals feel it personally.
Your résumé is no longer a file you hand over. It is whatever the web says about you, aggregated by a machine, for someone who asked a casual question and got a confident answer. If your LinkedIn says one thing, an old bio says another, and your firm's site has not been updated since the merger, the model does not weigh the evidence and pick the most flattering reading. It hedges, or it names someone whose story is coherent.
The same applies to a consultant, a contractor, a practice, a winery, a manufacturer. Anyone who has ever been checked out before being called.
What to do about it
In rough order of how much they matter:
Be consistent before you are clever. One business name, one address format, one phone number, everywhere. This is unglamorous and it outperforms almost everything else, because ambiguity is the single most common reason a model skips a business.
Be corroborated by things you do not own. Directories, review platforms, industry associations, local press. These are the sources assistants actually read when they answer a local question. Absence from them is not a gap in your marketing — it is a gap in the evidence a model has to work with.
Be structured. Schema markup is how you state facts in a form a machine cannot misread. It is not a ranking trick; it is the difference between a model inferring what you do and knowing it.
Answer the actual questions. Write the page that answers what someone types, in the words they type, and answer it in the first two sentences rather than the eighth paragraph. Models quote what is quotable.
Then measure it. Not your rankings — your presence. Ask the assistants the questions your customers ask and see whether your name comes back. Most businesses have never once checked, which is remarkable given what is at stake.
See where you stand. Our free audit reads your site the way an AI system does, then asks a live assistant real customer questions and shows you whether your business comes back — and which sources it used to decide.
Run the free audit →The window is still open
Most of your competitors have not thought about this at all. They are still optimising for a search results page that fewer people look at every month, and they will keep doing it until the phone stops ringing and they cannot work out why.
That is an advantage, and it is temporary. The businesses that get named in these answers over the next year will be the ones that were deliberate about it while everyone else was waiting to see whether it mattered.
It matters. The document you used to hand over is gone. Something else is answering for you now, and it is worth knowing what it says.