Getting cited by AI, one month in: I built a free tool instead of more blog posts

A month ago I wrote about asking ChatGPT to recommend an expert in my own category and watching it name four competitors and not me. I traced the answer to a directory I sit below, found that the lever to climb it was structurally closed to my business, and said I was going to stop chasing that one query and go win the other hundred.

That post ended with a plan. This one is the part that matters more. What did I actually do first.

I want to write it down honestly, because the answer surprised me. The first real move was not another blog post. It was a tool I gave away for free.

Why I didn't just write more

The obvious read of the last post is "write more content for the non-directory queries." That is true, and I am doing it. But a few weeks of looking at how AI actually picks sources made me change the order of operations.

The thing I had underweighted last time was brand mentions. Not my pages. Other people's pages talking about me. Per Ahrefs data from December 2025, brand mentions correlate with AI visibility about three times more strongly than backlinks. When I went looking for mine, I was close to invisible off my own site.

Here is the problem with fixing that by writing. More articles on my own domain do not create mentions on anyone else's. I can publish every day and still not show up in a single Reddit thread, YouTube video, or comparison article. The signal I was weakest on was the one my own blog cannot move.

So the question changed. It stopped being "what should I write" and became "what would give other people a reason to mention me without me asking."

What earns a mention

People link to two things without being asked. Original data, and free tools.

I will get to the data. The tool was the one I could ship fast, because it solved a problem I already understood better than almost anything else. Every client I take on starts the same way: they have a CRM or a spreadsheet full of years of messy history, and getting it cleanly into Attio is the boring, fiddly, error-prone part that scares people off the whole move.

So I built a free tool that does exactly that. You point it at an export from HubSpot, Pipedrive, Affinity, a spreadsheet, whatever you have, and it maps and imports the whole thing into Attio. No call with me required. No fee. You can run it yourself.

It took real work. It is its own project, built in phases, and it now handles the main CRMs and plain CSVs. I did not gate it behind a lead form. That was deliberate.

Why give it away

The math is the whole point.

A blog post I write is a page I own, talking about myself. It helps, but AI discounts it because everyone's marketing pages claim to be the best. A free tool is different. When someone uses it and it saves them a painful afternoon, they have a reason to mention it that has nothing to do with my marketing. They write the Reddit comment. They put it in the thread where someone asks how to move off HubSpot. They list it in the roundup of Attio resources. Those are the exact mentions I was missing, and I cannot manufacture them by writing. I can only earn them by being worth mentioning.

There is a second effect. The tool gives me a page that answers a real, high-intent query directly. Someone searching how to import a CSV into Attio does not want an essay. They want the thing that does it. Matching that intent with an actual tool, not a sales pitch, is the kind of page AI is comfortable pointing people to, because it resolves the question instead of selling against it.

And there is a third effect I care about as a founder, not a marketer. Giving away the scary part of the migration is the most honest version of my pitch. If you can do it yourself with my tool, great, I genuinely want you on Attio. If you get halfway in and decide you would rather someone owned the whole setup, you already know I am the person who built the thing that got you started.

What I am not claiming

I am writing this two days after the tool went out, so I am not going to pretend it changed my rankings. It has not had time to. I still sit below the directory cutoff for that one recommendation query, and I expect to for a while. That is fine. That was never the query I decided to win.

What I can say is that I now have something on the board that can earn the signal I was missing, instead of another page that cannot. The bet is simple. Useful free thing, given away with no gate, earns mentions from people who are not me, and those mentions are what AI weighs most. I will know in a few months whether the bet pays. I would rather be wrong about a tool people can use than right about an article nobody links to.

The order I wish I had known

If I were starting this over, I would do it in this order. Build the genuinely useful free thing first, because it is the only asset that earns mentions you did not have to beg for. Write the content that explains it second, structured so an AI can lift a clean paragraph. Chase the directory you can actually move last, if at all.

I had that order backwards for a year. I wrote first, owned everything I published, and wondered why the open web stayed quiet about me. The web does not talk about your writing. It talks about what your writing is about. So I built something worth talking about, and now the writing has something real to point at.

The next thing on the list is the data. I sit on a quiet pile of information about which tools teams leave when they move to Attio, and that is the other thing people link to without being asked. But that is the next post. This one was about the tool, and about the lesson that took me too long to learn: the best way to get cited is to be worth citing, and you cannot write your way to that. You have to build it.