Dosa batter is the most patient piece of product development I know. You grind rice and urad dal, leave the bowl somewhere warm, and then you do the hardest thing in any craft: you wait, and you don’t touch it.
I love dosas more than is probably reasonable, and somewhere along the way I realised that making them had quietly taught me most of what I believe about iteration.
You can’t rush it. You can only set the conditions.
Nobody has ever sped up fermentation by being more motivated. In Bengaluru’s warmth, a bowl on the counter is ready by morning. In a cold kitchen the same batter sulks for a day, which is why people in colder cities tuck it into an oven with only the light on. The batter hasn’t changed. The environment has.
Design research works the same way. You can’t force an insight out of a session by wanting it harder. You can set the conditions: the right people, the right task, a calm room, no leading questions. Then you wait and watch.
Change one thing at a time
If a batch comes out flat, the temptation is to adjust everything at once: more dal, less water, more salt, a different bowl. Then the next batch is better and you’ve learned nothing, because you don’t know which change did it.
The same applies to a design that isn’t performing. Change the label, the layout and the flow in one release and you’ve bought yourself a mystery. One variable per round is slower on paper and much faster in practice.
Judge by signals, not just the clock
Recipes say “ferment for eight hours.” The batter doesn’t read recipes. You look at it: has it risen, is it full of small bubbles, does it smell pleasantly sour rather than sharp? The clock is a guideline. The batter is the evidence.
I think about that every time a dashboard says everything is fine. Metrics tell you the number. Watching a person use the thing tells you why the number is what it is. I want both, and when they disagree I trust the person in front of me first.
The first dosa is for the pan
Anyone who cooks dosas knows the first one comes out wrong. The pan isn’t hot enough, or it’s too hot, or the batter sticks. Nobody is upset about it. It’s understood to be the test, and you throw it away without ceremony.
I wish design teams could be this relaxed about first versions. The first draft is the one that heats the pan. It’s allowed to tear. It’s there so the second one can be good.
Taste it, then change something
The feedback loop is slow, which is the cruel part. You adjust today and you find out tomorrow. So you learn to keep notes, to run small batches, and to take the result as information rather than a verdict. A thick dosa isn’t a personal failing. It’s a note for next time.
That last point is the one I’d pin above every designer’s desk, mine included: a bad result is data, not a verdict on you.
Everyone’s batter is right
Ask five households for their ratio of rice to dal and you’ll get five confident answers, and all five are correct for their kitchen, their grinder, their weather.
That’s the best argument for research I know. There isn’t one correct answer to bring in from outside. There’s only what works for the people in front of you, and the way to find it is to ask, watch, adjust, and ask again.
Now, if you’ll excuse me, I have a batter to check on.