Audience capture

Audience capture is the part that scares me the most about writing on the internet. I have the soul of an engineer¹: when I see a metric like engagement, I want to improve it. I like to make the number go up.

In fact, the whole environment of online writing is tailor-made to help you make engagement numbers go up. It’s trivial to find out how many people are reading my posts, and from what sources. Hundreds of link aggregators all over the internet have algorithms that will help my most engagement-friendly posts get views, and the comments are helpfully sorted by upvotes so I can read the “best”² pieces of feedback straight away. Many platforms are set up to financially reward you for engagement, whether that’s by enabling advertisement or by straight up paying you out.

Leaning into this environment is a huge mistake. Views and engagement are valuable to me, but not directly. They are proxy metrics for the things I actually care about:

• Having smart, informed people read and consider my ideas
• Getting feedback from readers who have concrete experience with the topics I write about
• Joining interesting public conversations about the state of the industry
• Convincing people to think like me on points I care deeply about
• Building a reputation as a careful thinker and a good writer³

It does not benefit me at all to have ten thousand people click through to one of my posts, scan the first sentence, then close the tab. It benefits me only a little⁴ to have a hundred people read the title of my post on Hacker News, decide not to click through, and leave their comments in the thread. The set of people who do those things — we could call them “clickers”, as opposed to “readers” — does not really overlap with the set of people who will actually read my posts and think about them. And yet in terms of metrics and engagement, they are indistinguishable.

As with any proxy metric, optimizing it too hard can cause it to diverge from the actual value it’s targeting. This leads to the scariest form of audience capture: being captured by an audience you don’t really even want. Suppose I get a lot of views when I write posts about how stupid it is to use microservices⁵, because it’s a hot topic. That gives me a reason to write more posts about microservices. The numbers keep climbing. However, I suspect the set of people who read my posts thoroughly — the audience I actually care about — don’t really want to read a bunch of the same screeds about that, so they drift away⁶. The people clicking don’t care because they’re not reading the posts anyway. So I’m actually failing, not succeeding: I’m replacing my audience of careful thinkers with a fake audience of clickers and sneerers.

This is not just a post about writing on the internet. In 2025, I wrote Trying to impress people you don’t respect about how succeeding as an engineer involves, well, impressing people who you may not necessarily respect. The core idea here is knowing what game you’re playing. Some games are proxy games, like getting views on posts or having great Jira ticket metrics. Some games are real games, like gaining the respect of people who you respect, or doing technical work you can be proud of, or trying to convince people of things that you believe are true.

The difference is in whether you care about the goal of the game, not in whether that goal is objectively valuable. If you really want to get promoted, then that’s a real game; if you just want to do solid technical work, getting promoted might be a proxy game that you play in order to get into a position to do that. But even though you don’t really care about it, it’s easy and tempting to win the proxy game at the expense of the real game: for instance, to crush Jira tickets because they’re in front of you instead of doing what will actually get you promoted, or to get views on the internet by playing into clickbait.

Making tradeoffs like this is sometimes fine. For instance, you might decide to impress people you don’t like so you can afford to send your kids to a good school. Selling out a little bit is good and healthy⁷. But do it with eyes open! The absolute worst thing you can do is to sell out by habit, in situations where it doesn’t even buy you anything you want.

There are a lot of posts that I really enjoyed writing that did not get much attention: typically posts where I get a bit more into the philosophy or history, or posts that are dry technical explainers. Some of my most popular posts are laments about a changing industry that are really just popular because they give people an outlet to share their worries too, not because of anything I wrote. I think it’d be a big mistake to lean into writing more posts like that and fewer posts like this. For one, if I want to just optimize for engagement, why write blogs at all? Why not post ragebait and pornography? But the main reason is that I’m writing for the 5% of readers that will read my posts, think about them, talk to other people about the ideas, email me with their comments, and so on.

Raw view counts help expose my writing to more of those people, but they don’t give me much signal on whether I’m appealing to them or not. For that I have to just rely on my taste, or on the thin signal I get back from the readers who do email and message me with thoughtful responses.