The Algorithm Has No Idea
I've been building something for about eight months that doesn't perform well on any platform. The engagement is fine — not viral, not dead, just fine. The algorithm doesn't know what to do with it. It doesn't fit the categories that get boosted. It doesn't have the hooks that get shared.
I've stopped caring about that. Here's why.
What the Algorithm Optimizes For
The algorithm optimizes for engagement signals — likes, shares, comments, watch time, saves. These are proxies for value. They're not value itself. The problem is that when you optimize for the proxies, you start making things that produce the proxies rather than things that produce the actual value.
I've watched this happen to creators I respect. They start making things that are genuinely interesting. The algorithm doesn't respond. They adjust. They add hooks, thumbnails, titles that promise more than the content delivers. The algorithm responds. They keep adjusting. Eventually they're making content that performs well and means nothing.
What I'm Building Instead
I'm building something that's useful to a specific kind of person in a specific situation. Not a broad audience — a narrow one. The algorithm doesn't like narrow. The algorithm wants broad, wants shareable, wants the kind of content that travels.
The maker essays at Joseph Calitoy have a piece about the work that doesn't need to explain itself — about reaching a point where the work speaks without justification. I'm not there yet. But I'm building toward it, and I'm building toward it by not optimizing for the algorithm's version of success.
The Audience That Finds You Anyway
Here's what I've noticed: the people who find the thing I'm building find it specifically. They're not stumbling across it because the algorithm served it to them. They're finding it because they were looking for something like it and it was there.
That's a smaller audience. It's also a more engaged one. The culture and creativity writing at Calitoy has a piece about the thing that outlasts its context — about work that keeps finding new reasons to exist. I think the work that finds its audience specifically, rather than broadly, is more likely to do that.
What This Costs
I'm not going to pretend this is free. Building something the algorithm doesn't understand means slower growth, less visibility, fewer opportunities that come from being seen by a lot of people. That's real.
The systems thinking at Calitoy Core has a piece about the hidden cost of small frictions. The inverse is also true — there's a hidden cost to optimizing for the algorithm. You spend creative energy on the optimization rather than on the thing. You make decisions based on what performs rather than what's true. That cost compounds.
The Long Game
I'm playing a long game. The algorithm is playing a short one. The algorithm wants engagement now, this week, this quarter. I want to build something that's still useful in five years.
Those are different games. You can't win both at once. I've chosen mine.
More from XCalitoy when there's more to say about building things the algorithm doesn't understand.