loader image
January 30, 2017 /
To Blend or Not to Blend

This post is an exploration of Blending: why businesses blend coffee, why I believe some roasters choose to blend, and why others don’t. I also suggest why blending has a bad rep within the Specialty Industry.

For a very long time “coffee” was a blend by default. Some businesses have their own namesake as a blend, many larger businesses have consistent year-round blends, and more recently, specialty businesses have been favouring the seasonal blend.

Blends are shrouded in mystery by many roasters. It’s a dark art, practised by legendary roast masters who claim it cannot be taught. They keep the percentages a secret, and some don’t even write the recipe down lest it be copied! I find this pretty ridiculous and frankly, masturbatory. Many roasters can’t roast a coffee with perfect consistency week to week. This makes the thought of copying a blend well enough to be a commercial threat rather farfetched.

There’s a few reasons why a coffee company will make a blend. Usually, it’s correlated to the size of the business; the economies of scale it affords and the consistency at volume it demands. I do acknowledge that some companies sit between or across some of the categories below. Just take this as a general guide to the economies of their main blend rather than the business as a whole.

Large Scale Roasters

Probably not focused on serving high grade specialty customers: Illy, Starbucks.

Large scale roasters need a consistent product in gargantuan volumes. To keep some semblance of consistency, they buy a constant stream of many coffees and blend them together to mitigate change. Seasonality is less of a concern because there are enough coffees and flexibility in the blend to buffer any seasonal changes in flavour. The downside of this method is the lack of transparency and character displayed by the blend components. To be honest, in a blend of 8 components, it really doesn’t matter how floral that Yirgacheffe is.

That said, it’s truly incredible just how consistent some of these blends are. The logistical and commercial demands placed upon the professionals developing them are mind boggling, and my hat goes off to them.

Medium Scale Roasters

Serving mostly specialty focused customers: Intelligentsia, Blue Bottle, Sensory Lab.

Medium scale roasters will generally create blends that are as consistent as possible without sacrificing the “voice” of the components or seasonality. This means they might choose 2-4 countries for a blend and stick to them throughout the year. If they’re smart, those countries will have two crops per year, or a widely dispersed crop time. This means the blend can contain fresh coffees all year round without changing flavour too drastically. It also provides them with some wiggle room to adjust the blend ratios without losing transparency of flavour. For example, at Sensory Lab and St Ali, we have a blend called Steadfast that is always Brazilian and Colombian. We utilise the main and fly crops of Colombia and the varied harvest times of Brazil to keep it incredibly consistent.

Small Scale Roasters

Almost definitely serving specialty customers: Heart Coffee, Square Mile, Market Lane Coffee.

Small scale roasters will, nearly by default, create blends that are quite variable and seasonal. A great number of Specialty roasters name their mainstay blend the “Seasonal Blend” and others will have an offering that includes a consistent blend, as above, alongside a seasonal blend that’s more variable. This rapidly changing blend happens for a few reasons depending on the business.

Some will explain that coffee is inherently seasonal and that they want to celebrate that by offering a constantly changing blend according to what’s the freshest. This is the best-case scenario. Others are just bowing to the pressure of small-volume coffee buying. Without many economies of scale or a strong cash flow, it’s in one’s favour to buy a lot of smaller lots more regularly (not to mention that the highest quality coffee almost always comes in small lots anyway). Buying a container of Brazil would sink many a small roaster, so the oft-changing blend is sometimes preferred.

A constantly changing seasonal blend is great, but makes for an inconsistent product. Some might think of the constantly changing Seasonal Blend as a carefully crafted experience – and it can be – but for a lot of roasters it’s a commercial necessity rather than a luxury. It’s often the case that a roaster’s blend will include the waste or “buffer” inventory of the business. As the most popular product the blend can be a great way to sell coffees that are getting long in the tooth or under-performing on their own. Although it may not always hurt the quality of the Blend (it might be a great coffee that just isn’t selling well!) it does damage the integrity and original intent of the business to sell one.

So that’s the why roasters blend from a rather commercial standpoint. But why do some roasters choose blending while others stay entirely away?

Pro Blending

Blending is often touted as a way to make a product greater than the sum of its parts. From this standpoint, blending allows coffees to hide their faults and show their strengths. I believe this is often just wishful thinking. If you blend one thin, acidic coffee, and one rich and sweet coffee, you don’t magically get a juicy rounded cup. You get a coffee that’s a bit acidic, a bit sweet and has a mouthfeel somewhere between thin and rich. Blending doesn’t just hide the negatives and bring out the positives, it puts everything together and dilutes those qualities according to the ratio.

Some go so far as to say that a single origin/estate/microlot coffee will never taste good on its own and must be blended to taste great (with the aggressive improvements towards coffee quality in recent years this opinion is certainly less popular). I definitely don’t agree with this, but I will happily admit that blends can – occasionally – taste significantly better than their components. This is rare, and involves a great deal of skill, but it’s far from impossible.

Anti Blending

Others have gone so far as to stop blending altogether; notably Tim Wendelboe in Oslo and The Barn in Berlin. I believe the prevailing argument here is that the coffees being bought are incredibly complex, unique and special, so blending them would only serve to diminish their qualities. I think this is a noble and just cause – their coffees are indeed special – but I’m not convinced that this is the best course of action for the industry at large.

I don’t believe the majority of coffee-drinking customers are quite ready to appreciate and understand why roasters have suddenly dropped blends from their offering. Blends still hold a special place in customer’s hearts; they’re the flag-bearer for the business; the gateway to the rest of your offering. Don’t get me wrong: I’m definitely a strong advocate for exposing customers to unique and special coffees. I’m just wary of how we go about that as an industry.

Tony Konecny, founder of Tonx Coffee (recently acquired by Blue Bottle Coffee) wrote a great piece on blends here. He wears the different hats of purist, cynic, master and optimist quite well. Check it out for another perspective on these matters.

The Real Dark Art of Blending

I can see why a lot of roasters and baristas don’t like blends. Quite often they’re a disappointing representation of the coffees within. You take a chocolaty Brazilian and a juicy Kenyan, put them together and you get a powdery, dry, sour blend. But why? Shouldn’t you get fruity chocolate?! No. Not unless you understand solubility and how it plays a part in blending

If you have ever tasted the components of a blend separately, and then together as a blend, you may have noticed that the blend tastes nothing like the best of its components. In fact, it tastes like the worst of them. This is because the coffees have differing solubilities. Please, allow me to explain.

Back in my post on Talking about Roasts I mentioned that a developed roast makes a coffee more soluble. In other words, it allows water to permeate its structure and dissolve its flavour. Unfortunately, solubility isn’t a yes/no situation. There are many shades of solubility, and every coffee is a little different.

Think of an aeropress with 6g of a very soluble coffee (A) and 6g of another coffee that’s less soluble (B). From the moment you add water, those two coffees will extract at different rates. Let’s stop the brew at different points and check in to see what those coffees will taste like.

2:00 – A will be powering towards a full extraction, beginning to produce some nice sweetness and roundness. B will be struggling to keep up, and is still very under-extracted. To combat that sourness you let the brew go on.

3:00 A is starting to taste a bit over extracted. B is only beginning to taste good, and needs more time.

4:00. A is well over-extracted and, finally, B is tasting great! … Damn.

At no point during that brew did both of the coffees taste their best. No matter how skilled you are as a barista, the brews you make of this blend will always be a compromise. There is, quite literally, nothing you can do to fix that. The best you can do is to stop the brew around 2:30-3:00 where neither coffee is aggressively under- or over-extracted.

[Some call this mix of roast flavour “melange” and strive for it as a desirable characteristic. Just like over and under-extraction being boring, generic and definitively ‘non-specialty’ flavours, this mix of roast flavour has less to do with the coffees used and more to do with the varying levels of extraction created because of it.]

I’m about to throw out a very personal opinion that many won’t agree with.

If a roaster isn’t fastidiously matching the solubility of every component in a blend it will never taste great. There’s nothing anyone can do to fix it. Once blended, varying solubilities of coffee will always be extracted differently and will taste muddled or incoherent. I really, really dislike blends with varying levels of solubility and extraction. I will not waver on this one. It’s game over for me.

When adjusting the ratio of coffees in a blend with varying solubilities, you’re not adjusting the intensity of each coffee’s flavour. You’re actually adjusting how disparate their extractions will be. A sobering thought.

I know of a few companies that measure and match the solubility of their blend components. It’s also really difficult to manipulate roasts according to that kind of information. I have a very strong suspicion that this is the root cause of why many specialty coffee professionals don’t like blends. It’s not the blending itself, it’s a result of blending coffees with differing solubilities.

When performing quality control on blends for Sensory Lab, I’ve been blown away by just how much better things are when the solubilities are matched. It’s night and day. When mismatched, the blend will taste jumbled and confused; it’ll be a big mix of under-, ideal- and over-extraction. When matched, every coffee can be easily identified. None of the components are hidden or diminished by the presence of the others, they just all exist peacefully side by side. It’s magical.

So if you’ve been sitting in the anti-blending camp, maybe give it another go. Talk to your roaster and understand a little more about the solubility of the components in a blend. Maybe ask for some of your blend components to taste individually, and play around with different ratios.

After all that talk about magical blends I can’t leave you hanging. Here are three examples of blends (geographically dispersed to avoid postal pains) that I know are definitely solubility matched. Maybe order some and give blending another chance. See if you can taste the individual components all at once, rather than a boring old melange!

Square Mile’s Sweet Shop is, continuously, a wonderful expression of seasonality and vibrancy. Right now it’s a mix of Aida Batlle’s Kilimanjaro (washed) and a Yirgacheffe.

Heart Coffee’s Stereo is a treat. I love the concept of 50:50 as a fixed blend ratio, and right now it has two Ethiopians; Schilcho and Reko. #rekosuave

Sensory Lab’s Seamless is on point right now. One of my all-time favourite Colombians is the majority, followed by a soft Panamanian and a sprinkle of rich Brazil.

Have an opinion about blends? Completely disagree with me? Great! Let’s chat below.

 

If you have found this useful and want to enjoy delicious coffee with the rest of the community – register for our monthly Superlatives coffee subscription. Or if you just want to keep up with every thing Barista Hustle – sign up to the Newsletter.

Coaching Calendar

Find a course with a BH Certified Coach

April 2020

Mon Tue Wed Thu Fri Sat Sun
1
  • Untitled
  • INA: Barista One
2
  • INA: Percolation
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
  • FR: PERCOLATION
25
  • FR: BARISTA ONE
26
27
28
29
30

News & Updates

Sign-up, Take part and keep in touch!

67
Leave a Reply

Please Login to comment
avatar
  Subscribe  
newest oldest most voted
Notify of
JohnR
Guest
JohnR

For all of us who learned to brew pour overs by adjusting our grind size to achieve a 3 minute per cup rate of flow, the implication here is that we’re probably not getting the best results this way. Is that correct?

Jan F-F
Guest
Jan F-F

Standard procedure is with a refractometer. Either get a very standardized brew prep like cupping (you need to filter out insolubles, as with any non paper filtered method) or use espresso. Measure the TDS and %EY of each component. All components in the blend should produce identical TDS.

Marshall Hance
Guest
Marshall Hance

As well. I don’t trust cupping bowls. I’d think just measuring the espresso itself is the most obvious and direct method of knowing how soluble a coffee heading towards espresso is… But what exactly others are doing is highly interesting to me.

Tim Chew
Guest
Tim Chew

I’m really keen to know what your method for testing the solubility of a coffee is. Would love to start experimenting on this myself!

SlowRain
Guest
SlowRain

Can someone help explain what “the main and fly crops of Colombia” means?

TrTruman Severson
Guest
TrTruman Severson

I had once spoken to Scott rao about the idea of matching solubility by blending green coffees and sealing them together in a bin or tub or otherwise. His thought process behind it was that given time, via osmosis, the coffees moisture contents’ would become nearly identical therefore making the two coffees much more similar in how they absorb heat. I don’t know if he had ever tried it. It seems to me that if this is true it might be a really good way to work around (read: cheat) The difficulty of solubility matching. Thoughts?

Jonathan White
Guest
Jonathan White

Interesting perspective and insight that I haven’t thought of. Thanks for the article!

While blending two or more coffees with different solubility in theory will result in a poorly extracted drink have you done any experimentation into blending two or more coffees after they have both been extracted at their ideal rate/time? For example two separate single origin espresso shots that normally wouldn’t extract together but may taste great when poured side-by-side then combined.

Jan F-F
Guest
Jan F-F

I brew to achieve a given ideal extraction yield for a given coffee. This varies based on brew prep and roast but is generally 20-23%. Some coffees produce the same EY with longer extraction times (e.g. some washed Ethiopian coffees.)

Grant Conine
Guest
Grant Conine

I would think that the difference of two coffees’ solubilities would depend not just upon the moisture content of the respective greens, but also on green density, size, and shape. I think the answer is a lot more complex than just matching moisture levels (though that is definitely a step in the right direction).

Adam M
Guest
Adam M

When solubility matching our blend we: -Dial in the blend (as espresso) until we hit our recipe (and confirm, obviously). -Pour shots of each component with the same grind setting and time then measure. -Adjust roast profiles until all components reach the same ext yld with the same parameters (yld not tds as they will be different concentrations). We feel this best emulates what will occur ‘in the basket’ – coffees will have differing solubilities but also differing flow rates at the same grind, this needs to be taken into account. Apart from the ‘night and day’ epiphany Matt describe,… Read more »

David Smith
Guest
David Smith

My method is to keep a consistent dry dose (20g) and adjust the grind on each separate blend component to produce my normale espresso in 27 seconds using the volumetrics. Take a sample of three different brews for each coffee and note down the average tds%. coffee A might be 9%tds B might be 10% C might be 11%. I’ll then reduce the development on coffee C and increase development on coffee A. The target being that all three coffees produce a 10%tds in 27 seconds at the same dose. Anyone have some knowledge about how the particle size effects… Read more »

David Smith
Guest
David Smith

They have two harvests a year. The main harvest is the big one, the fly crop is the crop that grows in between the yearly main crops.

JohnR
Guest
JohnR

Ideally, then, we should all be brewing with refractometers.

Gregory Levine
Guest
Gregory Levine

Matt, thanks for a very thought provoking explanation of blends. This material is great, and I’ve learned from every Hustle so far. My question is, what do you think of roasting components separately versus roasting components together?

Tim Chew
Guest
Tim Chew

My other question for Matt is, how would you go about picking coffees that have compatible solubility, prior to brewing?

Are there certain varietals or regions that go together?

Is it very dependant on roast degree as well?

Alex Bitsios-Esposito
Guest
Alex Bitsios-Esposito

Another excellent post Matt, thank you. I might add that even single origins – single estates – to my eyes, are technically blends, are they not? They are blends of the properties that they possess. I mean to say that we can further sort and grade coffees to even more unique micro lots. If we take Square Mile’s Sweetshop as an example, Aida Batlle’s Kilimanjaro is a blend of Bourbon and Kenia varietals. The Yirgacheffe will contain heirloom varietals – it may contain longberries, shortberries, peaberries; I cannot find information on the screen size of these coffees. My question is:… Read more »

Marshall Hance
Guest
Marshall Hance

This seems pretty smart…. although if flow rate is significantly different for the two components at a given grind setting, the results would be confounded.

Adam M
Guest
Adam M

How so?

Joshua Lindstorm
Guest
Joshua Lindstorm

What about pre blending the coffee before roast? How does this effect the grind consistence and being able to pull a nice shot? Talked with a freind the other week and they pre blend the espresso and he stated that it is a lot easier, quicker to dail in the espresso. thoughts?

Christopher Schaefer
Guest
Christopher Schaefer

I tried this years ago with a retail blend from a roastery I used to work at. The major result was a more uniform roast but the final cup didn’t display any glaringly positive or negative effects. Each of the components varied with respect to bean density and size. I had tested a handful of cases and varied their pre-mix time. The bets result, if I remember correctly, was a month-long age. I don’t believe, however, that the time in it took to achieve this provided any value to me the roaster or to the consumer. It just tied up… Read more »

Adam M
Guest
Adam M

Agreed – which is why the approach instead pushes you (through the loop) to: 20% in 30s and 20% in 30s – albeit with (more than likely) differing concentrations.

🙂

Marshall Hance
Guest
Marshall Hance

I’m totally with you on specificity. in fact, I don’t cup since that’s not how I intend my coffee to be served. That said, If one component is yielding 20% in 20sec, and the other is 20% in 40sec, and the blend is 50/50… the combination might yield 20% in 30sec, but actually be 25% and 15% of the components respectively. Of course all those numbers are extreme and unlikely, but this at least better illustrates the concern I’m seeing with your approach.

Adam M
Guest
Adam M

1. Yes there will be different brew times – it is accounted for in the procedure. 2. How would a slower shot (and thus one where the coffee has less contact with solvent) extract faster without being a significantly (i.e. user adjusted) different grind size? 3. If we will be using the coffee as espresso, it has to be tested as espresso, surely? The QC has to match the intended use to be valuable, especially when there are this many variables bouncing about. (Re control – many, many shots.) 4. If we adjust our grind to achieve the same flow… Read more »

Marshall Hance
Guest
Marshall Hance

For one, grind setting doesn’t necessarily imply grind size or distribution, so it’s entirely likely two coffees with similar solubility ground on the same setting may result in widely differing brew times. the slower flow will absolutely extract further. Therefore measurements of these two coffees would falsely suggest differing solubility. However, if both brew at the same flow rate, we can probably presume TDS measurements to be indicative of solubility. All that said, I’m thinking espresso may not be the most controlled test.

David Smith
Guest
David Smith

Just to chime in regarding inconsistent particle size in testing:
Use the same grinder?
Have your control blend in the hopper, dial it in to your brewing recipe, then remove that coffee, purge the grinder, then brew all your components as singles to the same dose without touching the grind size and record your results.

David Smith
Guest
David Smith

For me the short answer is simply; yes.
Long answer; Yes all those will give you a technically higher quality coffee and the results in the cup should exhibit a clarity of flavour rarely seen in. However a lot of those variables can add to the complexity of flavour as well (Ei varietal blending of the same estate coffee).

Matt Perger
Guest
Matt Perger

You, sir, just guessed my future post. Exactly.

Matt Perger
Guest
Matt Perger

Unless the two coffees are so identical that they’re in fact the same coffee – this will almost guarantee that the coffees will be roasted differently.

Matt Perger
Guest
Matt Perger

A good idea, but unfortunately moisture content is only a small piece of the puzzle. You would be able to get much much closer much more quickly by roasting and testing the components separately.

Matt Perger
Guest
Matt Perger

You should be able to roast ~98% of coffees to the same solubility somehow.

Matt Perger
Guest
Matt Perger

What he said!

Matt Perger
Guest
Matt Perger

Never together. Always separately.

Adam M
Guest
Adam M

Precisely 🙂

Lower retention grinders make it easier obviously.

Matt Perger
Guest
Matt Perger

@marshallhance:disqus @tim_chew:disqus @adam_marley:disqus Gents, good banter. If you grind different coffees at the same setting and make espressos, you’ll quickly find out which coffees are less soluble by the shot times. (faster = less soluble) Or you can do cupping bowls.

Matt Perger
Guest
Matt Perger

Yes! We call this post-extraction blending and it can be incredible. Jamie Thomson, one of our best Baristas did this for the Australian Brewers Cup last year with a Kenyan and Gesha.

Matt Perger
Guest
Matt Perger

Absolutely.

Mike Marquard
Guest
Mike Marquard

Matt – Such an awesome post! Thanks for this. I would love to see a post that goes into more detail about the roasting tools and adjustments you use to affect and test solubility…sort of a hybrid between this post and the “roasting” post. Do you also have “solubility” markers for your QC process (blends or SO) that is consistent across all coffees?

Jonolef
Guest
Jonolef

Seems like a worthy effort…though different varietals grown at different altitudes would still differ vastly in density and would therefore differ in their ability to absorb heat. Or not?

Jonolef
Guest
Jonolef

Two questions: 1. Am I correct in saying that there is the assumption that there is one given level of solubility where all coffees taste their best? That seems unlikely. And if each coffee requires a different level of roast/development/resultant solubility, then one might be able to match solubility perfectly across your chosen coffees, but the result is that one (or all) of the coffees will be roasted to a style that doesn’t bring out its/their best. 2. If you were not implying that all coffees taste their best at one give solubility level, then the next assumption would be… Read more »

IntlObserver
Guest
IntlObserver

Another thought-provoking piece. One additional thing I’d like to hear opinions on is how easy, and important, is it to maintain the integrity of the blend ratio? For example, the Sensory Lab Seamless blend is a 60/30/10 mix of three different coffees. Say you buy a 1kg bag. The beans will have been packed in an inherently abstract way, which will only be further disturbed as it is shipped from the roastery to a cafe and then ultimately emptied in to a grinder. One can only imagine that there would be a relatively large distribution of blend proportions for each… Read more »

Matt Perger
Guest
Matt Perger

If a coffee is less soluble it’s almost always less brittle as well. This makes for a “coarser” grind size. Best way to go without a refractometer.

SlowRain
Guest
SlowRain

Thank you, kind sir.

Damian Czyżewski
Guest
Damian Czyżewski

This always made me wonder. As above if as with many blends we have a less dense Brazilian natural/pulped natural rounder in a rounder shaped bean as a 70% base giving us body/mouthfeel and a lot denser coffee for example a heilroom African with a totally different bean shape as the remaining 30% giving us blend “balancing” flavour notes how do we know if we get anything near the balance the roaster planned in the 14-21g dose from a grind on demand grinder? I recently thought of trying to run a test by for example dying one part of the… Read more »

Marshall Hance
Guest
Marshall Hance

Whoa. How consistent is this heuristic? I mean, pulling shots is far quicker than dialing in the vst… let alone creating good samples. If true, this offers a rapid path for profiling for solubility.

El Cantante
Guest
El Cantante

I agree with a lot of this. I recently did a post on blends as well. My opinion was that blending was to achieve consistency (as you said). I also find that blends are roasted darker. That seems to me that roasters are trying to burn off the bad elements so you taste mostly “roast” and body, not much left of nuance.

Chester
Guest
Chester

Hi Matt, loving the Hustle and the post article conversations.
Being a little passionate about a single origin approach to coffee myself (some advantages too being a very small micro roastery, but not part of my reasoning behind my SO pursuits) Lately I have found myself experimenting with some single origin roast profile variation blending (max 3 profiles) with some great results in the cup.
Some thoughts on this as another approach would be appreciated.

Matt Perger
Guest
Matt Perger

Just have to cross your fingers and hope. There’s no way to nail it unless you’re weighing out each shot individually.

Valeri Tkatchenko
Guest
Valeri Tkatchenko

I think blending for espresso is important if done right it makes it somewhat easier to have a coffee that is much more balanced and crafted, I sent in a few single origins for espresso judging to AICA and they didn’t perform well but blended at a correct ratio it was judged as Champion espresso as a blend 🙂

Dan
Guest
Dan

I know roasting for solubility is a bit more complex, but to increase solubility of a coffee, would you increase the development time, total roast time, or final drop temperature, or a combination? Or do you have a particular method of doing this?

I roast a 3 part espresso blend for my shop, I think I’m very close to an ideal range, and have had success, but I still am not dead set on a method for increasing the solubility of a coffee.

Gunde Tyler
Guest
Gunde Tyler

dude, the Heart Reko was so fucking amazing. Everyone at the shop was floored by it. Gone now :/

Copyright © 2019 Barista Hustle, All Rights Reserved!

You have Successfully Subscribed!