My first sourdough starter died in about ten days, and I didn't understand why until I'd killed four more the same way. The sixth one is three months old now, feeds reliably, and doubles in about five hours at room temperature. This is what was actually different about it, as best I can reconstruct.
What Went Wrong the First Five Times
Starters one through three all followed the same arc: promising bubbles by day three, a nice rise by day five, and then a slow, quiet collapse into something that smelled like nail polish remover and never recovered. Starter four never got past a faint fizz. Starter five looked healthy right up until I left for a five-day trip and came back to a jar of grey liquid with a skin on top. Each failure felt like a different problem at the time. In hindsight they were mostly the same problem wearing different clothes.
The Six Changes That Actually Mattered
- Switched from tap water to filtered. Our tap water is chlorinated enough that I suspect it was quietly suppressing the culture in starters one and two — the fizz would appear and then stall in a way that looked a lot like the yeast never getting a real foothold.
- Went from all-purpose flour to a whole-wheat and all-purpose mix. Whole wheat carries more of the wild yeast and bacteria on the bran to begin with, and the difference in early activity was noticeable within two feedings.
- Fed on a strict twelve-hour schedule instead of "whenever I remembered." Starter five's grey-liquid death was almost certainly this — five days unfed at room temperature starves the culture past the point of recovery.
- Kept it somewhere consistently warm — on top of the refrigerator instead of a countertop that swung ten degrees between day and night. Wild yeast tolerates a range, but the temperature swings seemed to be exactly what pushed starters two and three toward the acetone smell, which is a sign of stressed yeast and overactive bacteria.
- Fed a smaller ratio of starter to flour and water — roughly 1:2:2 instead of 1:1:1 — which meant less competition for food between feedings and a more stable rise-and-fall cycle.
- Stopped discarding into the trash and started discarding into a jar in the fridge, which sounds irrelevant to the starter's health and isn't — it forced me to actually look at it every single day instead of only when I remembered to bake, which is when I caught the one near-miss where it had started separating and needed an extra feeding to recover.
The Loaf That Convinced Me It Had Worked
The real test wasn't the starter itself but the first loaf that used it without any commercial yeast as a backup, which had been my quiet habit through starters two and three — half wild starter, half a pinch of instant yeast, so I could never fully tell if the wild culture was doing any real work. The sixth starter, alone, produced a loaf with an open, irregular crumb and a crust that crackled audibly cooling on the counter, both things that never happened with the yeast-assisted loaves regardless of how active the starter looked in the jar.
What I'd Tell Someone on Starter Two
If your starter fizzes early and then stalls, it's very likely the water or the temperature, not your technique with the flour ratios — check those two boring variables before troubleshooting anything more interesting. I lost four starters to inconsistency in things I could have made consistent from day one, and the fifth to simple neglect. The sixth one didn't succeed because I got better at sourdough. It succeeded because I finally treated it like a living thing with actual requirements instead of a kitchen project I could get to eventually.