Obligation debt // Switching costs + downside risk

November 30, 2022 · writing

Simon Willison has 185 side projects. 185!!!! And given that he appears to be holding it more or less together, I was extremely interested in his recent talk: Coping strategies for the serial project hoarder. 

The talk’s opening is largely technical and tactical. But the last three slides just bowled me over, an absolute masterclass in side-project philosophy (images below are from Simon’s talk notes):

Guilt guilt guilt. How terrible to begin a project in earnest excitement and then feel trapped by your ongoing obligations toward it. And what is Simon’s answer?

Slight tangent: I was recently chatting with a couple founders about the the experience of someone signing up for your stuff and then immediately demanding more features. They suggested that people get what they’ve bought, at that particular moment in time, with no entitlement to any future improvements.

I didn’t fully agree, but it took a few days to figure out why.

The reason is switching costs + downside risk.

If I walk into a restaurant that doesn’t make me happy, then I can leave, no harm no foul. If I go so far as to order something, but I don’t like it, that’s still on me. No lock-in, low downside, no obligation. But if I hire a caterer for a large event and then realize, late in the day, that they refuse to serve anything vegetarian… Well, I think that’s different. Given that it’s extremely costly (in terms of finances, time, attention, stress, etc.) for me to switch caterers last minute, and that the downside is very large, I believe that they have do have some obligation – maybe not 100%, but certainly not 0% either – to accommodate my needs.

So rather than saying we don’t owe our customers anything beyond what they’ve currently bought, I think a fairer statement is to say that it depends on the idea: some ideas do carry future obligations; some ideas don’t. And user accounts – i.e., a container for storing your stuff – certainly pushes it toward the “yes obligation” direction.

We can run the heuristic in reverse: if you’re NOT up for ongoing obligations (which is true of side projects, passive income, and experiments), then you should avoid picking ideas that involve high switching costs and/or high downside risk for your users. Don’t collect their data. Don’t lock them in. Don’t even let them make an account!

A lot of founders will do it anyway, because switching costs boost retention, and downside risk (i.e., “being mission critical”) boosts pricing power. But that comes at the cost of either guilt (acknowledging that you screwed up and can’t keep up) or denial (pretending you didn’t even do anything wrong).

In either case, far better to pick ideas that fit your future constraints. The final piece:

Or in other words: wrap it up and tie a bow on it. Documentation is part of the product. So is customer success, onboarding, everything. 

Lots of the above that I need to get better at. Guilt is my primary emotion; it’s the water I swim in. Hearing Simon’s take, I’m optimistic. Yeah, I’m still going to have to deal with the shitty decisions I’ve already made. But I’m already dealing with that, so no biggie there. But now I’ve got a clearer heuristic for what I take on next, and a way to evaluate whether or not I’m going to be able to live up to it like I’d like to. And that’s exciting.

← all writing