Espresso extraction is a variable system with multiple interacting inputs and a single observable output: the liquid in the cup. Without a log, every adjustment is a guess with no reference point. With a log, every adjustment is an experiment with a hypothesis and a result.
Why Memory Fails the Home Barista
The grinder setting that produced a 28-second pull last Tuesday is not recoverable from memory on Saturday morning after three days off and a bag change. That number is somewhere between 12 and 15 on the stepless burr adjustment, you think, but you spent the first three shots of a fresh Guatemalan single origin finding it again because the previous dial-in was never written down. That is not a skill failure. It is a documentation failure.
The relationship between Grind setting, Brew Time, and the Input-to-Output ratio is the core of espresso extraction science, and none of it is intuitive in the early stages. Finer grind means longer brew time for the same target volume. Lower dose (Bean Amount) with the same grind setting produces faster extraction and lower yield. These variables move together in ways that are not predictable without data — they are predictable only after you have enough logged shots to see the pattern for a specific machine, a specific grinder, and a specific bean.
The Brew Time field records seconds. The Bean Amount records grams. The Volume records the output weight or volume. These three numbers together with the Grind setting are the complete extraction fingerprint for any given shot. When you pull a shot that tastes sharp, sour, and quick, and the log shows 22 seconds on a 17.5g dose at grind setting 13.2 producing 42g of output, you know exactly what happened: under-extracted, and the grind needs to go finer. When the next shot runs 31 seconds and tastes bitter and flat, the grind went too far and the log tells you exactly how far.
The Suggested New Grind Setting Field
The Suggested new grind setting field is the field that turns this log from a record into a tool. It is not the setting you used — it is the adjustment you intend to try next, recorded at the point of evaluation while the shot is still in front of you and the extraction characteristics are fresh. A shot that ran fast and sour at 13.2 gets a note of 12.8 in the suggested field. Next morning, that number is the starting point, not a guess.
This field is also the record of your reasoning. If you pulled a shot at 13.2, suggested 12.8, but the next morning forgot and went to 12.5, the log shows that the jump was larger than intended — and if the resulting shot is still acceptable, you have data suggesting the grind setting is not as sensitive to small changes as you assumed on that particular bean.
What the Photo Field Captures
A crema photograph taken immediately after a shot tells experienced eyes a significant amount about extraction quality. Pale, fast-dissipating crema on an espresso roast typically indicates a stale bean or under-extraction. Dark, oily, thin crema can indicate over-roasted beans or channelling. A thick, caramel-coloured crema with a fine, persistent mouse-tail pattern is what a dialled-in shot looks like on a bean with adequate freshness and appropriate grind.
Photographing each shot alongside the numerical log builds a visual library of extraction states that the notes field cannot fully replicate. Two months into logging a new machine, you will be able to look at a crema photograph and read the extraction before checking the numbers — because the photographs trained your eye in context with the parameters that produced them.