What people actually type to find a performer: what my own search data says

Of the search clicks my site got last quarter that arrived with a query attached, all but one were a performer's name or handle. Not a category. Not a definition. Not a question about the industry. A name, usually somebody else's, frequently misspelled.

That result reorganized how I think about this site, so I'm publishing the shape of it. I'm giving you ratios rather than raw totals, because the totals are my business and the ratios are the part that's useful to anybody else. Everything below is one small site's data over about three months. It is a real measurement of a small sample, not a law of the internet, and I'd rather hand you a real small number than a confident invented big one.

What did the data actually say?

Four findings, in the order they surprised me.

Names are effectively the entire query set. Of the clicks where the search term was attributable, 115 of 116 were a performer name or handle. Every other intent category I had built pages for combined — paperwork terms, definitions, payout questions, the bank-statement question — produced one click across more than ninety distinct queries. Not one click each. One click total.

Collaborator pages carry the site. The pages about the performers I shoot with are around three quarters of all clicks, off eleven pages. The blog, which is well over a hundred posts, is roughly an eighth. Per page, it isn't close.

One page is more than half of everything. A single collaborator's page earns more than the entire rest of the site put together. I'll come back to why that's a problem rather than a triumph.

Brand demand is essentially zero. Searches for my own name over the quarter: single digits, no clicks. The site ranks first for its own domain name and that query gets asked about once. Nobody is looking for me. They're looking for the people I work with, and finding me on the way.

Why are handles queries?

Because people search the string they saw, not the name a page is titled with. This is the finding I'd have most liked to know a year ago.

The queries that convert are things like lexymorgan0687 and missfortunenyc and bigbearforcuddles — social handles, typed verbatim into Google, digits and all. Someone sees a username in a comment or a repost, has no idea which platform it belongs to, and hands the whole string to a search engine. If your page doesn't contain that exact string, you are invisible to that search, no matter how well it covers the person.

Two corollaries fall straight out of that.

Misspellings convert. bbw lexi morgan with an i earns clicks at roughly the same position as the correct spelling. People type what they heard.

One modifier does most of the work. In my lane it's bbw, attached to a name over and over. A name plus a body-type word is a completely different, much softer search than either word alone, and it's the one that actually brings people in. I've written what that term means and where its relatives sit in the BBW explainer.

What does this mean if you run a creator site?

Put the strings people type on the page, and stop optimizing the pages nobody searches for.

Concretely, what the data supports:

• Publish every handle variant a person actually uses, with their permission, in the title, the description and the body. A handle you're guessing at is a liability; ask before you publish one.
• Spell the name the wrong ways too, somewhere honest, because the wrong spellings have their own demand.
• Treat a collaborator page as the product, not as a courtesy page hanging off the catalogue.
• Stop expecting definition and how-to posts to bring in buyers. Mine rank and earn nothing. They are useful for other reasons, and traffic is not the reason.

I want to be careful about one thing, because it's the part I got wrong first. I assumed pages earn clicks in proportion to how good they are, so the fix for a quiet page is a better page. Across my collaborator pages, page length, gallery size and internal links do not predict clicks at all. What predicts clicks is who is on the page. That is a genuinely uncomfortable finding, because it means the lever isn't craft, it's who you actually work with and how much demand their own name already has.

What's the risk in a result like this?

Concentration. If most of your traffic arrives through one person's name, you don't have a search strategy, you have a dependency.

That's my situation and I'd rather say it plainly than dress it up. One collaborator's page is more than half of everything, and there's no contract underneath that, just a scene we shot and a page she agreed to have up. The healthy target isn't more clicks on that page. It's no page above roughly a third, which you reach by adding people rather than by optimizing the one you've got.

The other risk is subtler. When the data says names win, the temptation is to chase names — to shoot with whoever has the biggest handle. That's a bad reason to book someone, and it produces exactly the scenes the audience can tell were booked for the wrong reason. How I pick collaborators is unchanged by any of this.

The one-line version

People search for people. If you're building an owned site in this industry, the pages that will earn are the ones about the humans in your work, spelled the way the audience actually types them, and almost none of the traffic will ever look for you by name.

That's not a discouraging finding. It's just a different job than the one most creator-SEO advice describes, and it's cheaper: fewer posts, better pages about real people, and the honesty to keep measuring after you've decided what you believe.

— Sly