The names “holiday blend” and “house blend” don’t exactly get the pulses of third-wave consumers racing. Blending has tremendous potential to mold or improve coffee flavor, but has an unfortunate reputation among many coffee lovers. 

Compared to “single origin” offerings, most blends are made up of cheaper, darker roasts meant to be “accessible” and paired milk and sugar. None of this is a commentary on what one should like; to each his or her own. I’m sure some stellar light, interesting “house blends” exist. 

“Everything is a blend”

At the farm level, cherry from various types of coffee trees may be blended and harvested together. At the dry mill, coffee in parchment or seed form, from numerous varieties or farms, may be blended at various steps to produce a lot with a single marketing name. A roaster may blend several coffees before or after roasting. A barista may combine multiple coffees to make a filter coffee or espresso. In a sense, almost everything is a blend, even coffees called “single origin.” 

The purposes of blending

Blending at the roastery level can serve many purposes. Roasters may blend coffees to maintain a certain flavor profile year-round, to try to “hide” an aging or otherwise disappointing lot of green coffee, or to attempt to save money while achieving a consistent flavor profile. Paradoxically, blending may be the easiest and most impactful way to improve coffee flavor after roasting. 

Blending can achieve an under-appreciated flavor synergy. One of the simplest ways to experience this is to blend a small amount of an intense, fruity natural coffee with a large amount of subtler washed coffee with lower-intensity fruit. While the natural coffee on its own may be too fruity or intense for some, it can spike the fruitiness of the blend just enough to create a “best of both worlds” effect. Blending can convert a “bug” into a “feature” by toning down its intensity. A roaster or barista may create almost any desired flavor profile through blending, but would likely struggle to achieve such flavor range if limited to using only one “single origin” coffee. 

Not many roasters or baristas bother to blend expensive, high-scoring coffees. After all, those coffees tend to be beautiful as is. But blending can sometimes improve even those coffees. At the end of our daily cuppings at Prodigal, Mark and I often spoon various proportions of coffee from different cupping bowls to create impromptu blends. More often than not, we prefer some of the blends to any of the individual components. The rare exception is when we have a nearly flawless, balanced coffee that is so much better than any other coffee on the table that it is nearly impossible to improve that special coffee. 

Pre-blend or post-blend? 

I recommend post-roast blending over pre-roast blending. The problem with blending before roasting is that beans of differing sizes and processes will develop at different rates. Pre-blending often makes ROR curves easier to manage, but may result in some beans being underdeveloped or a little too dark while other beans are developed beautifully. Pre-blending is generally not a good idea except when blending coffees of similar size and processing type. There is some folklore in Italy that says if you blend the green coffee for a few days before roasting a pre-blend, the beans’ moisture contents will homogenize, making them roast more harmoniously. I cannot verify if such moisture migration occurs, but would still avoid pre-blending due to variations among blend components’ bean sizes and processes. 

Roasting before blending allows a roaster to optimize the development of each blend component. Post-roast blending also allows the roaster or cupper to adjust blend ratios to optimize the roast batches on hand, which can be more effective than blending by formula if not all batches are on target. 

“Solubility matching,” the idea that one should blend only coffees of similar solubility, was once all the rage. The idea was that if (for example) a 50/50 blend consisted of, say, one coffee that would extract to 24% and another at 20% (using a given set of brewing parameters), each component would extract at a compromised, suboptimal level. However, this idea was never fully valid; the end result in the cup is an approximation of how the coffees would taste if brewed separately and then combined. Likewise, when blending two coffees that produce different particle size distributions (PSDs), the blend’s (PSD) will be the weighted average of each component’s PSD. 

How to Blend

There are no rules to blending, but I will offer a few recommendations: 

☞Choose post-roast blending over pre-blending. Post-roast blending offers more control and makes it easier to optimize the roast of each blend component. 

☞Use the spoon method: the best hack for deciding which coffees to blend, and in what ratios, is the spoon method: 

1. Set up a cupping bowl of each blend component candidate

2. Label the bottom of several empty cupping bowls with various blend ratios (such as 1:1, 2:1, etc)

3. When the coffees are cool, spoon various ratios of coffee from each bowl into an empty cup. For example, if you take three spoonfuls from cup A, two from cup B, and one from cup C, that would represent a 3:2:1 blend of those coffees. 

4. Make several such blends and have someone else shuffle the order of the cups

5. Taste the cups blindly to decide the optimal blend

☞Take chances. Try blends that don’t make sense or seem like they won’t work, and take the opportunity to think outside the box and learn about how coffees interact. 

☞Avoid blends in which one component makes up less than 15% of the blend. Even with adequate mixing, the ratio of blend components in each small dose of coffee is likely to vary. For example, in the 3:2:1 blend above, component C is 16% of the blend. If you brew using 20g doses of the blend, C may make up 10% of some doses and 20% of other doses. If a blend component were to make up only 10% of the total blend, some doses would have just a few percent of that component, making that component’s contribution to flavor almost invisible.