A founder needs someone to do exactly what you do. She’s not browsing casually. She has a real problem, a budget, and a decision to make. She opens ChatGPT and types the question her last three referrals would have answered on their own.
ChatGPT gives her an answer. Three names, maybe four, with a short reason attached to each. She skims the first two and books a call with one of them before lunch.
You weren’t in the answer. And you never find out.
No declined inquiry in your inbox. No bounce in your analytics. No unsubscribed email on a list you don’t have yet. The lead came and went without touching anything you can measure, and in a week you’ll wonder why things have gone quiet.
This isn’t a hypothetical. It’s happening across most professional service categories, and it’s been happening long enough that the pattern is clear. The question isn’t whether. It’s how often, and why.
Can you actually get into the AI answer?
Yes. The fix isn’t better work, it’s work the machine can read. Take Andre Obradovic, a men’s health coach. On his core client queries, not one engine named him. After about six weeks of citeability work, he’s now the first name ChatGPT, Perplexity, Google AI and Grok hand a client ready to hire, the client he would otherwise never have met, and he’s named as himself, with his real ICF and Primal Health credentials attached, not some flattened generic version. So if you’re not in the answer right now, that’s not a verdict on your work. It’s a gap you can close. The first step is to find out where you stand, then fix what’s making you illegible to the engines. The full case study shows exactly what changed.
Why can’t you see the lost lead?
Traditional lead loss leaves a trace. A bounce. A form abandonment. A prospect who went quiet after a proposal. Something to point at, diagnose, and fix.
This kind of loss leaves nothing.
When a potential client gets an AI recommendation and you’re not in it, there’s no record of the question ever being asked. No click, no visit, no open rate. The lead existed, briefly, in a question posed to a machine, and by the time the client lands anywhere you can track, the decision about who to contact first is already made.
This isn’t an analytics problem, it’s a structural one. The shortlist was built before you had any chance to be on it.
And the people who are on it aren’t necessarily better at the work. They’re just more legible to the machine that built it.
How common is AI-first client research already?
It’s tempting to treat this as a future problem. “AI search is still evolving.” “My clients aren’t using it that way yet.” It’s a comfortable position, and it puts off a decision you’ll have to make eventually anyway.
The timeline is shorter than it feels. ChatGPT passed 400 million weekly active users in early 2025. Perplexity has been growing as a research tool for exactly the kind of considered, complex decisions that lead to hiring a consultant or specialist. Google’s AI Overviews appear for a significant share of queries, sitting above the links your SEO work was meant to win.
The potential clients using AI for discovery are not a niche. They’re the competent, independent decision-makers who do their own research before they talk to anyone. Which is a fairly accurate description of the clients most founder-led businesses actually want. It’s also why ranking number one on Google no longer wins you the client the way it used to.
Some people in your category are already getting cited. The question is whether you’re one of them.
Why aren’t you in the AI answer if your work is good?
This is the part most people skip because it doesn’t feel as satisfying as an algorithm critique.
AI recommendation engines don’t evaluate the quality of your work. ChatGPT, Perplexity, Google AI Overviews and Grok can’t sit in on a client call, read your process documents, or assess whether your outcomes actually hold. They pull from what they can find, read, and quote across public sources, and they build a picture of you from that material.
If the picture is clear, they’ll recommend you. If it’s blurry, thin, or contradictory, they’ll go with someone who gave them more to work with.
For most founder-led businesses, the picture is thin. The expertise is real and the work is strong, but it lives in private client relationships, in tacit knowledge, in proposals that never become public content. What’s on the public record is often sparse, outdated, or pitched so generally that the engine has nothing much to grab onto.
That gap between your actual capability and what the machine can read is the citeability gap. It isn’t a quality problem, it’s a legibility one. And legibility is something you can design.
The three things an AI engine needs from you to recommend you are clarity (it can tell what you do and who for), consistency (your signals agree across sources), and quotability (it can lift a clean answer from your pages). If even one of those is missing, you’re at a disadvantage against someone who has all three, regardless of whether their work is better.
How AI engines decide who to recommend covers the mechanism in detail if you want the fuller picture.
What’s the first thing to do if AI doesn’t recommend you?
The worst response to this is a flurry of content. Publishing more without fixing legibility doesn’t close the gap, it just adds volume to a signal that’s already unclear.
Start by finding out what’s actually happening. Not guessing. Not assuming the best or the worst. Checking.
Run the prompts a potential client would run, in their language, looking for someone with your skills and context. Read the result without rationalising it. Note whose names come up, what the engine says about them, and whether your name appears at all.
The exact prompts and how to read the result are worth going through before you make any decisions.
Before you change a word of your content or update a single profile, run the check. Know your number.
The next potential client who asks an AI who to hire won’t tell you they asked. The only thing you control is whether your name is in the answer when they do. A Citeability Snapshot shows you the exact answer the engines give right now and the gap keeping you out of it.