Notes
I think the substack team is sincere about its mission and doing its best to manage the tensions of being a VC-funded platform while being writer-friendly and growing total readership. The biggest win has been avoiding advertising. But the “sovereign creator” archetype described by Hamish is a bit of dangerous mythologizing (and wanting to be more truly sovereign like Ben Thompson is a bit quixotic). But the other extreme of calling it “share cropping” etc as critics like to is equally inaccurate. Writers here are somewhere in between sovereign and sharecropper. I’d rather realistically acknowledge the main fronts of pragmatic compromise, which I think the team is actually managing well. The alignment is good but not perfect and cannot be. Here’s my take on the state of affairs:
Network effects are real now. The moat against competition that’s most reasonable to grow. I’d strongly support moves like bundling and tokens that allow readers more budget liquidity. This feels values aligned with writers as a group, if not individually. Happy to take a hit personally to grow the pie here for writers still growing/beginning (I’m at a plateau but a decently high one).
Substack now curates discovery which I’ve been wary of, first through spotlight mechanisms and new through notes that reward engagement farming. I think this is fine and reasonable but vulnerable to dangerous temptations. They have to double down on their most energetic revenue drivers (fine). But it’s nearing the knob limit where writers are starting to write for the discovery algorithms/curation tastes (not fine).
Subscriber list data and (in a janky way), Stripe billing relationships are “owned” by writers, including fairly good chargeback handling on our behalf. Good job. Not ideal, but best of a bad set of options. There is no way to do much better in fiat world. Crypto may eventually solve that.
Relationship tools like pausing subscriptions, gift subs, and pricing controls, are all great, but this is where writers need to keep an eagle eye because alignment is weakest.
Copyright is fraught and getting more so with AI training rights. I haven’t looked at the ToS in a while, but last I looked they have fairly extensive optionality to do more, especially on non-paywalled posts. I’m fine with this personally but some genres would face too much risk under this ToS. This will eventually to adverse selection of content.
Product growth vectors in short-form video and swipe-based low-friction graph curation — they don’t bother me at all personally as they do some people. Revenue being linked to longform core adequately balances those pressures for now. But if revenue gets length/form agnostic (eg monetizing notes), all bets are off. Shorter form and infinite scroll UX on app will eat the core.
Finally, perhaps my biggest concern is the gradual dominance of in-app reading and features over both email and web. The closed UX is eating the open protocol. What I’d really like I doubt will ever make sense for Substack to offer or most people to care about — content backup to my own website (not just domain) in the firm of some open-standard static site web framework. But data export is almost good enough. I’ve been thinking of setting up a pipeline for this. Periodic exports —> backup site. But it’s a heavy lift.
Overall, substack is about as good as Web 2.5 can get. But it’s ultimately in “don’t be evil” zone rather than “can’t be evil” zone as the onchain crowd says. I sense substack is mild-to-medium crypto hostile but long-term it feels like the inevitable direction. There are many more complex features and capabilities I’d like but providing them within the substack model will increase misalignment and lock-in. For example, I write lots of series, and on Wordpress I use a series plugin. But I’m wary of asking for a similar feature on substack because it creates more bespoke architectural lock-in and makes export data more idiosyncratic. Which means I’ll stick with the most commodity version of substack on offer. Case in point: My twitter archive is hard to turn into a static website because I used the threading and polls features extensively. So it’s just sitting in my backups
