Tracking sourdough and heritage grain schedules is a mess without a system
I started keeping detailed notes on my bread baking around 2014 because I was losing my mind. I would bake the same 70% hydration sourdough recipe three weeks in a row and get three completely different results, then have no idea which variable actually mattered. Was it the ambient humidity? The temperature of my starter when I mixed it? The brand of flour? Without writing things down, you are just guessing. After a few years of this, I settled on a structured logbook approach that actually stuck, and it is the closest thing I have found to getting reproducible results from wild-ferment bread. The system is simpler than people make it out to be. You need one page or one spreadsheet entry per bake, and you fill it out at five specific checkpoints during the process. Most people quit this habit because they treat it like a lab report and try to record everything. That does not work. You record the five checkpoints and nothing else. The first checkpoint is the morning you feed your starter before mixing. You write the date, the type of flour used for the feed, the water temperature, and the time elapsed before you mix. That last one matters more than it should. I spent two months thinking my loaves were dense because of my folding technique when the real problem was that I was mixing four hours after feeding instead of two, and the culture was already crashing. Writing the elapsed time made the pattern visible within a week.
Checkpoint two is the mix. You record hydration percentage, total flour weight, starter percentage, salt percentage, and the dough temperature right after mixing. Dough temperature at this stage is the single most important number you will write down. It determines how fast the bulk fermentation will go. If your dough comes out at 78 degrees Fahrenheit in summer and 82 in winter using the same room conditions, you are going to make the same mistake every time unless you track it. I ended up keeping a small thermometer clipped to my counter specifically for this step. Checkpoint three happens at the end of bulk fermentation. You do not need to time the bulk. You write the starting time, the ending time, the ambient kitchen temperature if you happen to measure it, and a one-word note on the dough texture. Stretchy. Weak. Tight. Bubbly. Aggressive. That is it. The one-word texture note has saved me more bad loaves than any calculation. You look back at three months of data and notice that every loaf tagged "weak" on a Tuesday had a bulk that was twelve minutes shorter than your norm, and you start asking why Tuesdays are different. It was because my starter was fed Monday evening instead of Monday morning, and the second feed was missing from my logs at the time. Checkpoint four is the shape and proof. Record proofing method, proofing time, and final dough temperature if you take it. I stopped measuring final dough temperature after a year because it turned out to be redundant with the bulk ending temperature for my setup. Redundant data is still data you do not need. Removing it is what kept the habit alive for me.
Checkpoint five is the bake and the result. Oven temperature, steam method, bake time, end crust color on a scale of one to five, crumb openness on a scale of one to five, and taste on a scale of one to five. The crumb and taste scales feel subjective until you realize your personal bias against a darker crust is consistent. Once you have thirty entries, you can cross-reference crust color with taste score and find that you rate a four-crust loaf two points lower than a three-crust loaf even when the flavor is better. That is useful data about yourself, not the bread. There is no app that does this well because the format is intentionally sparse. I tried three different baking journal apps and two spreadsheet templates before going back to a physical notebook. The reason is speed. If filling out the log takes more than forty seconds, you will skip entries during busy weekends, and the system breaks. A pocket notebook and a pen on the counter meets the speed requirement. I use a Moleskine that already has a grid layout, and I just write the numbers in order from top to bottom. No decoration. No diagrams. Just the five checkpoints. One edge case that almost killed my logging habit was a batch of spelt sourdough where the hydration jumped from 72% to 81% between when I wrote it down and when I actually mixed it because I misread my own handwriting. The resulting loaf was a pancake. I spent a week trying to figure out what went wrong before I realized the number was smudged. After that, I started using a fine liner pen instead of a ballpoint, and I circle every percentage number I write. It adds three seconds to each entry and eliminated an entire class of errors from my data. You will not believe how many of my early inconsistencies traced back to unreadable handwriting rather than actual baking variables.
Get the Full Details
The data becomes useful on two timescales. Short term, you review the last ten entries before your next bake to calibrate. If your last three dough temperatures were all two degrees higher than usual, you adjust your water temperature or mixing time accordingly. Long term, you review once a quarter to find patterns. The quarterly review is where you spot things like "my loaves on days when I feed my starter after 7 PM always have a weaker crumb" or "I bake better when the kitchen is above 70 degrees regardless of season." There are real limitations to this approach that nobody talks about. The system only works if you are the one doing the baking. If you hand your notebook to someone else and ask them to follow your notes, it fails immediately because they do not share your reference points for words like "stretchy" or "aggressive." The one-word texture descriptions are personal shorthand that means nothing to anyone else. I learned this when I tried to give my partner a week's worth of my logs to bake solo while I was away, and she produced four inconsistent loaves before giving up. She was not doing anything wrong. She just could not map her perception of dough behavior onto my labels. Another limitation is that this system does not help you learn new recipes. It helps you refine the ones you already know. If you are experimenting with a new grain blend or a new starter regime, the logbook will give you data, but it will not tell you what to change. You still need basic bread science knowledge to interpret the numbers. The logbook is a mirror, not a teacher. Beginners often treat it as a shortcut to good bread without putting in the reading first, and then they fill pages of data with no way to make sense of it.
If you want something more structured than a notebook, a simple Google Sheet works fine and has the advantage of being searchable. You can set up columns for each checkpoint and sort by dough temperature or hydration later. The tradeoff is that pulling up a phone or laptop mid-bake is slower than glancing at paper, and you are more likely to skip entries when the sheet feels like homework. I have seen people maintain paper logs for years while their spreadsheet siblings collect dust after six weeks. The Vintage Bread Making Logbook is not a product you download. It is a practice. But if you are looking for a starting template, the structure I described is free to copy. Five checkpoints, five lines of data each, one entry per bake. That is all of it. Fill it out consistently for ninety days and you will know more about your own baking than most people learn in two years of occasional trial and error. Do it for a year and you will have a personal reference library that no recipe blog can replace.
