So there's two things here: TILs and projects. I'm running a dev writers' retreat next week and happen to be thinking about this, want to suggest more:
- Data & News: This is a lot of work but collect data for yourself and publish it for others and others will find that useful. This is the strategy I am adopting for AI stuff https://lspace.swyx.io/p/open-source-ai. Dan Luu does this a lot to back up his work https://danluu.com/ The cutting edge of data is news, and you can see this on Gergely's substack (https://blog.pragmaticengineer.com/) but you're basically a journalist at that point
- Overviews: What your part of the dev ecosystem looks like to you. It'll never be perfect, it can't be. Some folks on HN will tear you to shreds for missing something obvious. But it will be immensely helpful to people just behind you in experience. See: How I write backends (https://news.ycombinator.com/item?id=22106482), The evolution of the data engineer (https://news.ycombinator.com/item?id=33317126), and my AWS vs Cloudflare post (https://news.ycombinator.com/item?id=28903982). Overviews become "Insights" when you can successfully reframe something in a new light that is useful for people to understand what is going on.
- Caching: Anything - ANYTHING - you look up more frequently than 1x a year (For concepts, I have a Three Strikes Rule https://www.swyx.io/three-strikes, For lists, I have persistent lists https://www.swyx.io/fave-podcasts and https://www.swyx.io/new-mac-setup). This could be a TIL, but often is not; the more accurate description is treating your blog as a "local cache" of stuff you always use, with exact instructions and copy pastable code that fits for you and is verified to work by you. The TAM is of course way smaller but this is how stuff like https://www.swyx.io/download-twitter-spaces ranks on Google immediately because of course there are way more people googling for the same things.
in short - try to identify the most useful parts of your experience and make it legible to others. your earnest and consistent effort elegantly answers "why should I read this" to the reader as well.
Writing about data seems reliably popular. There’s even a company that just writes viral blog posts for clients about their data (Pricenomics).
Whenever you run an experiment or gather data, consider whether it might be shareable as a post. It’s a win-win because people love learning the truth in numbers about a topic, and at the same time you get to have a chance at a decently popular post about your obscure thing.
Say you’re a toilet paper manufacturer. Probably nobody wants to read your company blog, but if you had a post with real data about how much TP is wasted by hotels not using rolls completely for example, that might just break through.
- Superlatives: An interesting story about the first/last/best/worst experience when you did X. You see a lot of these in Rachelbythebay's work (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...) . Mtlynch focuses more on documenting his journey semi-realtime (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...)
- Data & News: This is a lot of work but collect data for yourself and publish it for others and others will find that useful. This is the strategy I am adopting for AI stuff https://lspace.swyx.io/p/open-source-ai. Dan Luu does this a lot to back up his work https://danluu.com/ The cutting edge of data is news, and you can see this on Gergely's substack (https://blog.pragmaticengineer.com/) but you're basically a journalist at that point
- How $POPULAR_THING Works: This is Fly.io's (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...) and Alex Xu's strategy https://blog.bytebytego.com/
- Overviews: What your part of the dev ecosystem looks like to you. It'll never be perfect, it can't be. Some folks on HN will tear you to shreds for missing something obvious. But it will be immensely helpful to people just behind you in experience. See: How I write backends (https://news.ycombinator.com/item?id=22106482), The evolution of the data engineer (https://news.ycombinator.com/item?id=33317126), and my AWS vs Cloudflare post (https://news.ycombinator.com/item?id=28903982). Overviews become "Insights" when you can successfully reframe something in a new light that is useful for people to understand what is going on.
- Predictions/Wishlists: What is missing from your part of the ecosystem. The complement of the above. You can also invert it into a complaint post, like https://circleci.com/blog/its-the-future/ or https://hackernoon.com/how-it-feels-to-learn-javascript-in-2... but those are less constructive if entertaining
- Principles: Timeless things, a superset of the previous two points. See rauchg (Nextjs cocreator - https://news.ycombinator.com/item?id=8559519) and dan abramov (react core team - https://news.ycombinator.com/item?id=18792373
- Caching: Anything - ANYTHING - you look up more frequently than 1x a year (For concepts, I have a Three Strikes Rule https://www.swyx.io/three-strikes, For lists, I have persistent lists https://www.swyx.io/fave-podcasts and https://www.swyx.io/new-mac-setup). This could be a TIL, but often is not; the more accurate description is treating your blog as a "local cache" of stuff you always use, with exact instructions and copy pastable code that fits for you and is verified to work by you. The TAM is of course way smaller but this is how stuff like https://www.swyx.io/download-twitter-spaces ranks on Google immediately because of course there are way more people googling for the same things.
in short - try to identify the most useful parts of your experience and make it legible to others. your earnest and consistent effort elegantly answers "why should I read this" to the reader as well.