Build a coffee blend by tasting, not folklore
A blend is useful when combining coffees gives you something you prefer: a different balance, a steadier espresso, or a cup that stays recognisable as stock changes.
There is no required architecture and no magic ratio. ‘Base, accent, bridge’ can prompt ideas, but it is not a law. Taste the coffees, state the aim, test clearly different ratios, and keep the brew method constant.
Reviewed 9 September 2026
Begin with a purpose
Write one sentence: perhaps a forgiving chocolate-led espresso, more fruit in a filter blend, or a cup that works with milk. Taste every candidate alone using the intended brew method. Seller notes suggest candidates; your roast and brew reveal what each contributes.
Two coffees are enough to learn from. Add a third only to solve a named problem. Complexity is not a quality measure, and extra components make cause and effect harder to see.
Pre-blend or post-blend
Pre-blending combines green coffees before roasting. It saves steps and can work when the components tolerate the same treatment. Commercial roasters use it successfully; it is not inherently inferior. Different size, density, moisture, processing, or desired end points can make it harder to control.
Post-blending roasts each coffee separately. At home it is often easier for experiments because you can taste each component and test several ratios without roasting again. It costs time and may leave small amounts. Compare both methods rather than assuming one must win.
Run a ratio test
Blend by roasted weight. Prepare three clearly separated two-coffee samples, such as 80/20, 65/35, and 50/50. These are test points, not recipes. Brew with the same dose, water, grinder, method, and rest interval; taste blind if practical.
Refine the most promising area only if you can perceive the difference. There is no universal rule that five percentage points is detectable and two is not. Sensitivity depends on the coffees, brew, and taster.
- Foundation
- The component supplying most of the sweetness, body, or familiar character.
- Contrast
- A component added for a clearly perceived difference such as fruit, florals, acidity, spice, or bitterness.
- Adjustment
- An optional third coffee used to solve a named problem. If its job is unclear, leave it out.
Scale and record it
A 240 g blend at 65/35 uses 156 g and 84 g. Available roasted stock may set the maximum batch. Calculate a post-roast blend from roasted stock, not the green amount you began with.
Record the lot, roast date and description, ratio, rest time, brew method, and result. A ratio alone is not a recipe; the same coffees can change after a different roast or several more days of rest.
Common questions
Should I blend before or after roasting?
Both can work. Post-roast blending gives an experimenter more control and easier ratio tests. Pre-blending simplifies workflow and can succeed when coffees suit the same treatment.
How many coffees should a blend contain?
Start with two because their contributions are easier to identify. Add a third only when you can state the problem it should solve.
What is a good first ratio?
There is no best ratio. Try clearly different samples such as 80/20, 65/35, and 50/50, then refine the region you prefer.
Sources and further reading
These references support the technical claims above. Practical recommendations are identified as starting points because coffees and equipment differ.
Put this batch to work.
Percolate connects green stock, roast observations, blends, and tastings. Use it when you want the last cup beside you while you plan the next roast.
Start roasting