Building for an audience of one

I built Forma to manage my own training, and I'm its only user. I wanted a way to see my training that the apps I had weren't giving me, so I made it, and I've been living inside it ever since.

Most of what you do as a product designer is learning to see through other people's eyes. You sit with users, you take on the way someone else reads a problem, someone whose week looks nothing like yours, and over the years that turns into a real skill. The designers I admire are quietly very good at it. With Forma I've ended up doing the inverse, because the person I'm designing for is me.

That's the part that's supposed to make a designer uneasy, and I understand why. You're trained to keep your own taste at a slight distance, to treat what you feel as something to check rather than trust. But I'd say you never really stop designing from what you see and feel anyway. Here it's just the whole of it, turned up as far as it goes.

I'm not claiming to stand in for some segment of users. I'm the one person I know better than anyone, and I happen to have a real need for what the thing does.

There's a lot that's good about that. I don't have to go looking for someone to test on, because I'm already the test. I can ship a change at night and feel it on a run the next morning, in my real training rather than a script I set up to check it. When something's off I usually notice before a chart would point it out. So I keep adjusting as I go, holding onto what works, dropping what doesn't, and living with each new version until it shows me the next thing to fix.

Some of it goes deeper than feel, though. Heading into my build for the Toronto Marathon in May, and now the TCS Waterfront this fall, I was rotating three pairs of shoes without a real sense of how to split them across the work. I could sort them roughly by instinct, but the gear stats gave me the actual shape of it. They showed me how my pace and effort shift between each pair, and which one to reach for on a given day.

The EVO SLs drifted toward my shorter, faster sessions. The Boston 13s settled in as daily trainers for easy and long runs. The Gel Nimbus 27s took recovery runs and long walks. Forma even tells me when to reorder, more or less. My Boston 13s are around 350km now, the foam's starting to feel a little flat, and since they'll be spent near 500 I put the replacement in before I'd normally have thought to.

But there's a catch. I'm inclined to like what I built, and I'm inclined to believe I'm improving, and the trouble is that both of those lean the same way. It's a little like grading my own homework. Left alone, I'll give myself a good mark on both counts, the product and the progress, and not always because I've earned it.

So the work is being willing to answer no. As the builder, that means asking whether a feature actually holds up, not whether I'm fond of it because it's mine. As the athlete, whether I'm genuinely fitter, not whether the app told me so in words I enjoyed reading.

The place I feel that most is rest. The steadiest read in Forma is cumulative load over time, whether I'm building, holding, or quietly running myself into the ground. Left to instinct I'll push through a flat week, because I want the next session more than I want the day off, and instinct is glad to back me up. The load curve won't, and I built it partly to argue with me.

I'll be honest, though, that I'm still learning to trust it. It isn't the full picture yet, because it can't see my sleep or my resting heart rate, the recovery side my Garmin tracks and Forma doesn't. So I'm still reading the curve against how my body actually feels and tuning from there. Some days I take its word. Some days I overrule it and find out later which of us was right.

The real test is in the fall, when I race at the TCS Waterfront again, a full year after my first marathon run on the route. The training either delivers or it doesn't. Until then I'm doing both jobs at once, building the thing and being the only person it's for.