Most people who write to us expect to hear that something is wrong. They assume there is an old article somewhere, or a review that got out of hand, or a forum thread they have forgotten about, and they want us to find it and deal with it.

So the first thing we do is search their name the way an investor or a journalist would, and then we ask an AI assistant who this person is, and in most cases we come back with the same finding.

There is nothing damaging out there, because there is almost nothing out there at all.

That tends to land as good news at first, and we understand why, since a clean slate sounds like the best possible starting point.

But in any transaction where the stakes are high, a blank result is a problem in its own right, and in our experience it costs people more than a mild negative ever does. This article is about why that happens and what a real fix looks like.

Where the search actually happens

Before a serious meeting, someone runs a check on you, and that check has three parts which almost everyone now performs in the same order.

This is no longer an informal habit on the investor side either, since Vcheck Global, which runs background investigations for financial institutions, reports that seven out of ten top-performing venture funds now conduct deep checks on founders before committing, and that requests for those checks grew 26% in a single year.

The AI layer is the newest part and the one people underestimate.

Research from Omnibound on how business decisions get made found that 94% of decision-makers used an AI tool during their most recent major purchase or partnership process, that generative AI chatbots have become the single largest influence on early shortlists at 17.1%, ahead of a company's own website, and that 29% of buyers now start their research in an AI tool before they open Google at all.

SEO Sherpa's 2026 report on AI search adds the scale behind those habits, with ChatGPT alone at 900 million weekly users this year, up from 400 million a year earlier.

The uncomfortable part is when this check happens. The same research shows that around 60% of a buying decision forms before the buyer ever contacts the other side, and that 95% of the people who eventually win were already on the shortlist by the time the first conversation took place.

What nothing looks like from the other side

When the person doing the search finds nothing, they experience it differently from how you imagine. You picture a neutral gap, a simple absence, the digital equivalent of an unlisted number.

They picture something closer to a question mark, and question marks make people cautious.

Part of this comes from how AI tools handle an empty record. Google, when it finds little, shows you little, and an experienced reader can shrug that off.

An AI assistant, when it finds little, still produces an answer, and that answer usually says the model has no reliable information about this person. Those words read like a verdict, even though they only describe a lack of data, and the person on the other side rarely stops to make that distinction.

The other part is friction. When there is coverage to read, the investor, the board or the counterparty can confirm most of your story before the meeting, so the meeting itself becomes a conversation between people who already trust the basics.

When there is nothing to read, every claim you make has to carry its own weight in the room, which means you spend the first half of the meeting proving that you exist and the second half trying to get to the point.

Serious people can do that, and many do, but they walk in with a deficit they had no idea was there, while someone with a thinner track record and a fuller online presence walks in with a head start.

Why a mild negative is often easier than a blank

This surprises people, so it is worth being direct about it. A single old article that puts you in a bad light is a defined problem with a defined edge, and a reader can weigh it against everything else they find.

An investor who sees one difficult story from 2019 next to a decade of substantive coverage will usually read it in proportion, because the rest of the record gives them the context to do so.

A blank record offers no context at all. There is nothing to weigh the doubt against, so the doubt fills the whole frame.

We have seen founders with two profitable exits lose a first meeting to competitors with weaker numbers, and when we later asked what happened, the answer came down to the fact that the other person was easier to verify.

Being legible has become part of being credible, and legibility is something you can build on purpose.

Why the default profiles fail to solve it

The natural reaction, once someone sees the gap, is to fill it with the profiles that are already lying around.

It feels like progress, and we understand the instinct, but it changes almost nothing about what the three checks return, for a few reasons that are worth understanding.

The first reason is that these profiles are self-authored, and everyone doing the search knows it. A biography you wrote about yourself carries about as much weight as a reference you wrote about yourself, so the reader treats it as a starting claim to be checked rather than as evidence.

What people are looking for is third-party confirmation, meaning a publication that decided you were worth an interview, a journalist who chose to profile you, an editor who ran your commentary because you knew the subject. That kind of coverage is difficult to fake, which is exactly why it counts.

The second reason is about how AI systems decide what to believe. Research on how language models select their sources, summarized in the AI Thinker Lab guide to Generative Engine Optimization, shows that the authority of the publishing domain is the strongest single predictor of whether an AI will use a page in its answer, and data from LLMrefs puts a number on it: roughly 65% of the top thousand pages ChatGPT draws on come from domains with an authority rating above 80.

A personal website sits far below that line. A LinkedIn page is a directory entry that the model treats as one weak signal among millions.

The third reason is consistency. SEO Sherpa's data shows that brands present across four or more independent platforms are 2.8 times as likely to show up in a ChatGPT answer as brands concentrated in one or two, because the model gains confidence when several unrelated sources agree, and the same report finds that 83% of the sources AI tools rely on sit outside Google's top ten results, which means ranking well on Google and being picked up by an AI have become two separate problems.

A single profile, however polished, gives the model nothing to cross-reference and solves neither of them.

What a real fix looks like

The honest answer is that closing the gap takes time and takes editorial work rather than profile maintenance.

What actually changes the result in all three places is coverage that other people wrote about you, published in outlets that search engines and AI systems already trust, and added steadily over time so that it looks the way a real track record looks. In practice that means a proper interview in a publication your audience already reads, a profile that explains what you built and why it mattered, and expert commentary on the questions your industry is arguing about.

The rhythm matters as much as the content. Ten placements in a single month look like a campaign, and both readers and algorithms have learned to discount campaigns.

Four or five strong pieces spread across a year look like a person who has been consistently worth writing about, which is what you want the record to say.

This is why we plan cases in months rather than weeks, and why we tell people that the best time to begin is well before the raise, the board search or the sale, since a record that already exists when someone goes looking is far more convincing than one that appears the week before.

The wider market has come to the same conclusion. Mordor Intelligence values the online reputation management market at $7.75 billion in 2026, and inside it the largest and fastest-growing segment, at over 16% a year, is earned media, meaning coverage that a publication chose to run. People who pay for this kind of work have learned that this is the part which actually changes what a search returns.

The good news, and it is real good news, is that a blank record is the easiest kind to fix. There is nothing to displace and nothing to correct.

Where to start

If you are unsure where you stand, the test takes two minutes and costs nothing.

If what comes back is thin, vague or ends in a shrug, you now know the actual problem, and it is a solvable one.

We are happy to run that check with you and give you a straight read on what it would take to change it. Tell us where you are and where you want to be, and within 48 hours you will hear back from a strategist with an honest assessment of what is realistic, how long it takes and what it costs.

Sources