An espresso recipe tells you how much coffee you used, how much liquid you collected, and how long the shot took. But those numbers do not tell you how the liquid tasted at different points during brewing.

A salami shot makes those changes easier to explore. Instead of collecting the entire espresso in one cup, you collect successive portions in separate cups and taste them side by side.

It is a useful learning exercise—not a shortcut to finding one “perfect” section of espresso.

What Is a Salami Shot?

A salami shot divides one continuous espresso into several samples. You move from one cup to the next while the machine keeps brewing, preserving the order in which the liquid leaves the portafilter.

Three cups are enough for a first attempt. More cups give you smaller samples, but also require more switching and make the workflow harder to manage. Clive Coffee demonstrates a three-cup version of this exercise. Clive Coffee’s salami shot demonstration

The purpose is to notice differences between early, middle, and late portions—not to assume that each cup contains a separate category of flavour compounds.

Why “Sour, Sweet, Bitter” Is Too Simple

You may hear espresso described as extracting acidity first, sweetness next, and bitterness last. That can be a memorable teaching aid, but it is not a literal map of what happens inside the puck.

Research collecting successive espresso fractions found that the concentrations of the measured components—including caffeine, trigonelline, and a chlorogenic acid—were highest in the first fraction and declined as more liquid was collected. Total dissolved solids followed a similar downward pattern under the studied conditions. This was chemical analysis, not a sensory rule assigning one taste to each fraction. Espresso extraction kinetics study, Foods, 2023

Your cups still contain overlapping mixtures. A bitter first sample does not contradict the exercise, and a pleasant final sample does not mean you collected it incorrectly.

Prepare a Familiar Baseline

Before dividing a shot, brew and taste one normal espresso using your usual recipe.

Record the coffee, dry dose, final beverage weight, grind setting, and shot time. Note when your timer starts: pump activation and first liquid appearing are different reference points.

For example, your baseline might use 18 grams of coffee to produce 36 grams of espresso. Those numbers are an illustration, not a recommended recipe for every coffee.

Keep the same coffee, dose, grind setting, and machine settings for the tasting exercise. Otherwise, you will be comparing more than collection stages.

A Practical Three-Cup Exercise

Here is a suggested starting workflow.

First, arrange three stable, heat-safe cups within easy reach. Label them early, middle, and late, and weigh each empty cup.

Choose two switching points that fit your normal shot. For example, if liquid usually flows for around 25 seconds after the first drop, you could collect the first eight seconds in cup one, the next eight seconds in cup two, and the remainder in cup three.

These intervals are only convenient starting points. Do not extend the espresso simply to fill all three cups.

Keep brewing uninterrupted while changing cups. For this exercise, measure the collection intervals from the first drop, and record that convention clearly.

If changing cups interferes with your scale’s automatic timer or tare functions, use an uninterrupted manual timer and follow the equipment’s operating instructions. If the setup is crowded or requires reaching through hot liquid, skip the cup-switching exercise and compare separately brewed whole shots instead.

Do not remove the portafilter during brewing.

Measure the Actual Output

Changing cups creates a weighing complication: the number displayed under the current cup is not automatically the total espresso collected across all three.

For a simple first attempt, stop at your familiar programmed endpoint or usual timed endpoint. Treat the resulting yield as approximate until you weigh the samples.

Once the cups are cool enough to handle, weigh each filled cup and subtract its empty weight. Add the three liquid weights to find the actual total yield.

Equal collection times do not necessarily produce equal amounts of espresso. If the flow accelerates, a later eight-second sample may contain more liquid than an earlier one.

That matters when interpreting the tasting. The final cup may represent a larger share of your complete drink than its position in a three-cup lineup suggests.

Taste Without Forcing a Pattern

Let the samples cool to a comfortable, reasonably comparable drinking temperature. Stir each cup before taking small sips.

Smaller portions can cool faster, so this is not a perfectly controlled comparison. Treat it as a practical tasting exercise.

Start with simple questions:

• Which sample tastes most intense?

• Where is acidity most noticeable?

• Does any portion show sweetness or a more pleasant balance?

• Which feels fuller or thinner?

• Does any sample leave an unpleasant lingering or drying finish?

You do not need elaborate tasting notes. “Sharp and intense,” “rounder,” or “thin with a dry finish” can be more useful than trying to identify a specific fruit.

Write down what you actually taste, even when it differs from the expected pattern.

Different Coffees Can Behave Differently

In Clive Coffee’s comparison, the final portion of one light-roasted coffee tasted mild and balanced, while the final portion of a darker coffee tasted rough and bitter.

That limited comparison illustrates why the last cup should not automatically be labelled undesirable. It does not establish a universal cutoff for light or dark roasts. Clive Coffee’s tasting results

Repeat the exercise when you change coffee rather than carrying one set of conclusions into every new bag.

Turn Observations Into a Recipe Test

The most useful result is a question you can test with a complete espresso.

If the final sample has an unpleasant finish, try a modestly shorter beverage yield while keeping dose and grind unchanged. Then compare that whole shot with your baseline.

If the complete baseline tastes sharply acidic and the later sample seems pleasant, a slightly longer yield may be worth testing.

These are hypotheses, not diagnoses. A late sample can taste mild partly because it is less concentrated. That does not prove the complete espresso needs more of it.

To explore strength separately, Barista Hustle describes a different exercise: brewing several whole espressos at different yields, then diluting them to approximately comparable strengths. That is not a salami shot, but it highlights why concentration can complicate flavour comparisons. Barista Hustle’s extraction tasting guide

Always Confirm With a Whole Shot

Do not assume that the best-tasting isolated sample is the best espresso recipe.

Likewise, combining the leftovers after tasting does not perfectly recreate the original drink. You have removed different amounts from the samples, and they have cooled separately.

Brew a fresh, complete shot to confirm any change. Stir it, taste it as it cools, and repeat before deciding that the adjustment is an improvement.

What a Salami Shot Cannot Tell You

This exercise cannot measure extraction percentage, identify exactly where channeling occurred, or reveal pressure inside the puck.

It also cannot prove that a particular cup is “under-extracted” or “over-extracted” simply because it tastes sour or bitter.

Its value is more practical: it helps you notice changes that are difficult to distinguish when every portion is mixed together.

Make Your Measurements Useful

A coffee scale and timer help connect those observations to a repeatable recipe. Record your dose, collection boundaries, individual sample weights, and actual total yield.

Whether you use a WeighMaster or another brewing scale, the important part is consistent measurement—not making every cup fit a predetermined flavour chart.

The goal is not to isolate a perfect slice of espresso. It is to understand your coffee better, then use that understanding to make a better complete cup.