Retention isn't just the amount of grounds left behind

Retention is often referred to as coffee “stuck” in a grinder, but using a single final number makes it easy to miss the real nature of the phenomenon. There can be grounds sitting in the grinding chamber, grounds clinging to the exit path, coffee exchanged between batches, and grounds released late when the machine keeps running. So retention should be understood as a dynamic system characteristic, not just the number of grams left after a single grind. For quality-control purposes, you need to know not only how much grounds is held back, but also how old grounds get released into a new batch.

Mass balance and dose deviation

The simplest way to see the problem is a mass balance. If you feed in a defined mass of beans and get less out the other end, the difference is a signal worth investigating. But that difference isn't automatically a fixed retention amount: part may still be inside the machine, and another part may come out on the next grind. Barista Hustle built a specific protocol for measuring the amount of coffee remaining in a grinder, in which the machine is cleaned to an accessible level and the grounds collected and weighed. The value of this protocol lies in clearly defining the measurement endpoint, rather than mixing retained grounds with a cleaning purge.

Data worth recording

Retention can throw off dose measurement in two ways. First, if you weigh the input beans but don't control how much grounds actually reaches the person brewing, the dose ratio can be off. Second, grounds from a previous batch can carry into the next, changing grounds age and blend composition for a period. This matters especially when changing coffee type, roast level, or machine settings. A machine with low average retention but high grounds exchange between two batches can still create an operational problem that a single retention number can't describe.

Separating retention from cleaning purge

A minimum dataset for mapping retention should include input mass, output mass right after the machine stops, the amount recovered during a cleaning procedure, measurement timing, machine settings, bean type, roast level, and machine condition. If possible, also record temperature and electrostatic conditions, since grinding research shows electric charge can affect grounds retention and behavior. Recording the time between batches is also useful since some machines can release grounds slowly when vibrated or run again.

Comparing machines correctly

Measuring retention without confusing it with cleaning purge requires a clear definition. If the goal is measuring the amount left after one grind, stop at a defined state and recover the remainder. If the goal is knowing how much old grounds enters the next batch, you need a suitable bean type or marker and then track the carryover portion. The two measurements answer two different questions. Combining them into one weighing produces a number that's hard to interpret and easily misused when comparing machines.

Finally, grinders shouldn't be ranked by retention alone. Grinding-mechanism design, exit-path geometry, static electricity, bean type, moisture, operating speed, and usage habits can all change the result. A reliable comparison procedure must keep these conditions stable and publish the measurement method. Retention then becomes a controllable variable that helps explain dose deviation and grounds-exchange ability between batches, rather than an advertising number detached from real conditions.

A retention map should be built on at least two data series: the amount of grounds left in the machine, and the amount of old grounds appearing in the following batch. The first series suits physical recovery measurement; the second needs a way to identify the source of the grounds. If you only measure the first series, you may miss cases where a machine retains little grounds but releases a significant portion right at the start of a new batch. That's why retention and carryover should be viewed as related but not identical quantities.

When changing coffee type, measuring carryover becomes even more valuable. A small amount of old grounds may be insignificant by mass but meaningful if the new batch is very small or requires a completely different flavor profile. So besides absolute grams, the exchange amount can be converted to a percentage of the new dose. This reporting method helps compare different batch sizes without turning the percentage into a universal standard for every machine.

Retention can also be expressed as a function of time after the machine stops. If grounds keep falling out over several seconds or minutes, the moment of weighing the output will affect the result. So “output” needs to be defined as right when the grounds stream stops, after a fixed waiting period, or after a defined cleaning-purge procedure. Only with the same definition can data be compared across days.

A good control system should also track retention after cleaning and after maintenance. If the baseline changes, old data shouldn't be mixed directly with new data. Instead, the intervention timing should be flagged and the normal range of variation rebuilt. This management approach turns retention into an operational indicator with traceability, rather than a fixed number attached to a machine's name.

References

• Méndez Harper et al. (2024), Strategies to mitigate electrostatic charging during coffee grinding, iScience 27, 110639, DOI 10.1016/j.isci.2024.110639.
• Barista Hustle (2018), A BH Protocol for Measuring Grind Retention.
• The Craft and Science of Coffee (2017), The Grind—Particles and Particularities.