Your former employer’s leadership page still lists you, years after you left, because nobody took it down. A conference page from 2019 and a podcast episode description carry their own versions of your career, and you wrote none of them.
None of it was malicious. An organizer needed forty words by Friday and wrote them from your LinkedIn headline. A host summarized your career from a fifteen-minute pre-interview call. A directory scraped something once and has never looked at it again.
This one makes me angry on behalf of the people it happens to. Somebody spends decades building a career, and the version of them a stranger finds first was typed in five minutes by somebody who never checked it. Nobody meant any harm, and the person it describes is the only one who pays for it.
That mattered less when a person looking you up read a page and formed an impression. It matters more now, because the thing forming the impression is frequently a machine assembling an answer from everything it can find, weighting a 2019 conference bio much the same as anything you published last month.
You don’t get a vote in that process. What you can do is understand how it picks, and then produce enough correct material that the picking goes your way.
Where do the wrong versions of you come from?
Other people wrote them, and most of them were doing you a favor at the time.
A conference organizer needed forty words by Friday and wrote them from your LinkedIn headline. A podcast host summarized your career from a pre-interview call. A former employer still has a leadership page with your old title on it. A directory scraped something once and has never updated.
Every one of those is a public record of who you are, and every one was written fast by somebody with partial information and no reason to check.
They don’t go away. Old pages outlive the events that produced them, and a machine reading about you now weighs a 2019 conference bio the same as anything you published last week.
Why does the wrong version keep winning?
Because it’s short, consistent and repeated. Bad information travels well when it’s easy to copy. One inaccurate forty-word description gets reused across a dozen event pages, each one linking to the others, and the repetition becomes the signal.
Your own site probably has the accurate version. It also probably has it once, in longer form, in a place designed for humans reading a page.
Old pages have age on their side, and age reads as reliability to a system that can’t tell the difference between established and stale.
So the picture that gets assembled about you is built from many small consistent wrong sources and one large right one. That’s not a fight your accurate version wins by being accurate. I think this is among the most unfair things about being findable now. The careful, accurate account loses to the sloppy one because the sloppy one got copied, and the systems doing the copying have no idea who was right.
What does a book do that a bio cannot?
It settles the questions the short versions leave open.
A forty-word bio can say you’re an operations leader with twenty years of experience. It can’t say what you believe about how operations should be run, what you decided in 2011 that you’d decide differently now, or which problems you’ve solved that most people in your field have never faced.
A book answers all of that at a length nobody can compress into an inaccuracy.
More usefully, it produces records that other people create about you. A catalog entry. A library holding. A citation in somebody else’s work. Those aren’t marketing material and they’re not written by you. That’s exactly what makes them count.
How do the machines decide which version is you?
By looking for agreement across sources that don’t appear to be coordinated.
A system assembling an answer about a person is trying to work out which claims are safe. One page saying something is a claim. Six unrelated places saying the same thing is a fact, as far as the system is concerned.
That’s why the wrong bio wins when it’s everywhere and the right one exists in a single place.
It’s also the route out. A book creates entries in registries, catalogs and retailers that weren’t written by you and don’t link to each other, and they agree because they’re all describing the same object.
Why are the profiles you control more important than they look?
Because they’re the sources a machine treats as authoritative about you, and they’re the ones you can fix this afternoon.
A system trying to establish facts about a person weights self-published profiles differently from third-party mentions, and it looks for consistency between them. Two profiles with different job titles for the same period is a contradiction, and a contradiction makes every claim in both less usable.
So identical beats good. The same job titles, the same date ranges, the same organization names spelled the same way.
Most people have four or five of these and have never read them side by side. I tell people to spend an hour doing exactly that before they touch anything else in this subject, because it’s the highest-return hour available and it costs nothing but attention.
AI really, really wants consistency. It wants the same short bio and the same facts everywhere you show up. I had to do a lot of work retrofitting my own profiles, because I’d been at this for 15 or 20 years and had old numbers sitting out there. I had to go find each one and figure out how to fix it.
How do you fix an outdated bio on a page you do not control?
Some you can, by asking, and the rest you outweigh. Conference organizers, podcast hosts and former employers will usually correct a page when somebody emails and asks politely. Almost nobody asks, and I can’t blame the organizers for that part. A wrong bio that sits for a decade because its subject never sent a two-line email is a problem you can end this week.
Where that fails, the answer is producing enough correct, independent records that the wrong ones stop being the consensus.
A book does that better than anything else available, because it generates entries in catalogs, retailers and registries that other organizations created and that agree with each other because they describe the same object.
How long does the wrong version take to fade?
Longer than you’d like, and the direction outweighs the speed.
Old pages have age on their side, and age reads as reliability to a system that can’t tell the difference between established and stale. A 2019 conference bio has been sitting there being consistent for years.
What changes the picture is volume and recency together. New correct records, created steadily, from sources that aren’t you. This is a slow instrument. It’s also close to permanent once it works. That’s the opposite of everything else in your marketing.
What should you fix first in your online bio?
The things you can change today, before anything else. Your own site should carry one description of you, complete, in one place, and every other page on the site should point at it instead of paraphrasing. Paraphrases drift and drift is what creates the second version. Then the profiles you control. Same facts, same job titles, same dates. Not similar. Same.
Then ask the organizations that have you wrong. Most will fix a page if somebody asks, and almost nobody asks. That’s an afternoon of email against a problem that will otherwise outlive you.
What does an inaccurate bio cost you?
The gap between what you’ve done and what somebody can confirm.
A stranger deciding whether to hire, book or introduce you spends a couple of minutes checking. They’re not reading carefully. They’re looking for whether the picture holds together.
When the versions disagree, nothing dramatic happens. There’s no rejection and no conversation about it. They move on to somebody whose record was easier to verify, and you never learn that you were considered.
That’s the whole cost, and I think it’s the worst kind, because you never see it. It runs every time somebody checks you, and it grows while you do nothing about it.
My take
Somebody else’s version of your career is out there, and machines are reading it. Email the conference organizers and podcast hosts and ask them to fix their pages. It takes two lines and almost nobody does it. For the pages you can’t change, outweigh them with correct records from sources other than you. A book does that better than anything. It’s slow, so start this week.
What does a complete professional bio contain?
More than a bio and less than a career history. The facts a stranger needs to place you. What you do now, who for, and at what level. Where you did it before, with real organization names and real dates. What you’re known for, stated precisely enough that somebody could check it.
Then the things a short bio never has room for. What you’ve built, published or been responsible for that’s a name. Anything a third party can verify without taking your word for it.
I ask my ghostwriting clients to start thinking about their author bio during the project, because it goes at the end of the book. They can have me interview them for it or send me one or two pages to work from. A summary of that bio gets reused on the Amazon book page and in promotion, so it won’t get thrown away after one use.
I’d build that page before touching anything else here. A stranger will make up their mind about you in two minutes whether you’ve prepared or not, and I’d much prefer they do it from something you wrote carefully over forty words somebody typed on a Friday in 2019.
Machines assemble your history from whatever they find. The Day Your Website Died covers optimizing for AI search so the answers about you come from pages you control.
