July 5, 2021
It is OKR season at my day job and so it is OKR season in Buttondown as well. I think I am actually making a pretty good go of it, this time around:
Two meaty projects (maybe a month each?), two smaller (maybe one-to-two-weeks each?) projects.
If you weren't around six months ago, I did a similar undertaking at the end of last year, with the ambit of it being a year-long plan and not simply a quarterly one:
It is sort of fun to re-read this. A couple notes:
- This was a good plan, and I mostly executed on it! The biggest exception was with that first big item around onboarding & conversion funnel work, which simply has not mattered — Buttondown has been growing too quickly to sweat about optimization.
- I massively underestimate how productive I can be when I'm focused on a line of work.
- The two habits I call out later in the document — avoiding "grazing work" and trying to be more metrics-driven — have really fallen by the wayside over the past quarter. That is honestly not giving me that much angst. I might start caring more about metrics at some point down the line, but it just doesn't seem important so long as I'm focusing on the topline metrics (engagement, MRR, churn, top of funnel).
So, one of the two things that are going to be top-line goals for this quarter is the docs site and I have managed to turn it into as fun an enterprise as possible. Behold!
Okay, there's not much to look at yet. But it feels fun to build, and gratifying. This is going to sit on top of Next.JS, Vercel, and MDX — a triptych of tools that I've been itching to play around with for a while now. Three obvious questions that I had to answer for myself as I built this out:
- Why not stick with Notion? This is an easy one. Notion is annoying to work in; it's difficult to refactor; it's bad for SEO; it's visually incongruous with the rest of the site; it feels amateurish; it is slow.
- Why not use something off-the-shelf, like Readme.io? I honestly wanted to go with this route, but the ecosystem for "arbitrary documentation tools" is surprisingly sparse! There are two buckets I found: the first bucket is "readme.io et al", and the styling & IA options seemed lackluster; the second bucket is "component of overall customer success tool like ZenDesk or HelpScout", which is not an ecosystem I want to get suckered into.
- Why not just build on top of the existing Django app instead of creating a new repository and codebase? A few reasons none of them super defensible: this seems more fun; I want to mess around with
nextjs; I want to open-source the documentation.
My hope is, as seen from the first screenshot of my "OKRs" (I know, these technically aren't OKRs, they're much more ship-based than outcome-based, but come on), to have scaffolding out pretty quickly and then migrate a bunch of the existing content to it. (Of course, this happens to blissfully delay the inevitable slog of writing the documentation, of which, yeah, I know, no amount of fun engineering or documentation harness building will ever replace the value.)
All of this feels really good. I don't like feeling like I'm drifting with Buttondown, and I definitely felt like that the past three weeks — which is a large reason why I squandered my time on silly projects like dependency cleaning or migrating to pytest. This feels good and concrete and just the right level of ambitious.
Here's hoping I have more documentation screenshots to share next week!