These are the results of the survey I conducted over the last twoÂ weeks. I had a great response to the surveyÂ – aroundÂ 1,500 replies. I needed to clean up a little bit of the data (which meant discarding quite a lot of weird responses), but there was still a lot to work with.
I hoped that the data would give some insight into global, as well as local trends. I was also curious to know if there was much correlation between home consumers and professionals, in their tastes and styles.
A caveat fairy early on: This is notÂ a truly good overview of how espresso is being served around the world. This is a review of how people who read this blog brew espresso in cafes and at home. I think this should be considered when looking at the results.
One final note: Brew ratios. I’m going to talk about relationship between ground coffee dose toÂ liquid espresso weightÂ as a ratio i.e. 1:2 or 1:1.5. I’m not going to use 50% or 66% etc, because these are percentages and not ratios. (I know that seems obvious but we have a habit of using brew ratio for both and it doesn’t make sense in my head).
Who is the fussiest?
Letâ€™s start with comparing the average commercial recipe with the average domestic one. Theyâ€™re actually very, very similar.
|User||Dose In||Dose Out||Ratio||Brew Time Range|
|Home user||18.3g||36.4g||1:2||24s – 36s|
|Commercial||19.0g||37.5g||1:1.98||25s – 34s|
The brew recipes and ratios are very, very similar. The only difference is that the acceptable brew time window is actually wider for home users. Interesting though, that the styles of espresso were pretty uniform on average. However, home users had higher standard deviations when it came to the brew recipe (essentially there is more diversity there), but lower standard deviations when it came to brewing time (there is more agreement on the window of acceptability).
The ristretto is dead
Not only is a 1:2 ratio, extremely common, and the global average for respondants, but the percentage of cafes and home users brewing ristretto style espresso is really very low.
This perhaps get amplified when I explain that I am categorising anything from 1:1.5 and below as a ristretto. (Which perhaps a very different definition than I might have used back when a 1:1 ratio was not uncommon).
Commercially only 11.8% of respondants are pulling shorter espressos/ristrettos, and about 14.5% of domestic users (perhaps an argument of consumer preference being slightly disconnected from commercial offerings).
I suppose I should try and explain away this data, and the trend in brew ratios so letâ€™s have a look at the distribution in cafes:
So, generally espresso is being brewed at a relatively high brew ratio and espressos being served are somewhatÂ dilute. While we can’t know people extractions, if we presume a reasonable level of extraction happening (which is quite likely with higher brew ratios) then lots of espressos out there around 9% strength. (For reference, many of us were previously chasing 12% and a brew ratio of around 1.5 until relatively recently – myself included.)
While some would smugly claim that the Italians were right all along (as an Italian espresso is often closer to a 1:2 ratio or more) the doses of coffee being used are a long way from Italyâ€™s traditional 7g per espresso. Also, the raw coffees being used are likely extremely different too.
I think a few different factors are likely at play here:
- Increased use of extraction measurement (and people chasing higher extractions)
- More experimentation with grinders and recipes (the EK43 thing etc)
- The internet echo chamber
More and more people are pulling shots in a similar way. Iâ€™ll come back to this point in the conclusions.
A Travellers Guide to Espresso in Different Cities
One of the fun things to look at was how different cities have different espresso styles. I’m torn how to present the data here. I looked at 20 different cities (that had large enough numbers of respondents to do a little data analysis). I don’t just want to dump the table here, so I thought I’d highlight a few different interpretations for some cities. Let’s start with my home town:
London: 18.5g In | 34.6g Out | 26s – 35s
London is interesting to me because it had the lowest levels of variationÂ across its recipes. Essentially the espresso style here was the most homogenous of any of the cities I looked at. Whether oneÂ interprets that as strong consensus or an embarrassing lack of diversity is a question probably better answered by consumers in London than by me. What also surprised me about London is that the average brew ratio put it in the top 5 cities for shortest/strongest espressos. The only cities pulling shots with lower brew ratios were San Francisco, Seoul, Moscow and Philadelphia.
New York City: Â 18.7G In | 35.5G OutÂ | 24SÂ – 33S
New York City was incredibly similar to London in many ways. Both have a very low percentage of respondents pulling ristrettos (about 4.5% each), but both also have relatively low diversity in recipes in the surveys. New York City is pulling slightly larger shots, slightly quicker but the averages are really very similar.
Sydney:Â 20.6G In | 47.2G OutÂ | 26SÂ – 35S
It was tempting to lump the main fourÂ Australian cities together (Sydney, Melbourne, Perth & Brisbane). What wasn’t surprising (going by stereotype) was that these were the only 4 cities with an average ground coffee dose higher than 20g. (Oh to be a roaster/supplier in Australia…)
I chose Sydney here, because it was something of a surprise to me. For a long time the stereotype of Australian espresso was a big dose in, and a relatively short shot. (Ah, the double rizzbanger…) This is clearly not the case anymore. Perhaps as a reaction to this particular stereotype, Sydney now pulls the largest and most dilute shots of any city. This larger brew ratio (1:2.3 here) is mirrored elsewhere. Perth is at 1:2, Melbourne and Brisbane both at 1:2.2.
One thing to note: home users in Australia aren’t pulling espresso the same way. The ground coffee dose is a little lower (about 19.5g) and the shots are smaller (1:1.8 is the average ratio). However, I don’t think we can confuse or conflate domestic espresso brewing preferences with consumer preferences.
Moscow: 18.3G In | 33.1G OutÂ | 22SÂ – 27S
Moscow was interesting, not only because the only place with a lower average ground coffee dose was Montreal (you didn’t see that coming…) but because the shots are pulled much quicker there than anywhere else. Pretty much every city is pulling shots for 30+ seconds. (Stockholm is the closest, with a similar ground coffee doseÂ and a brew time of 23 – 31 seconds.
I’d be interested to know what is driving this. Is it something in the water that aids extraction? Is it a general preference for shots brewed quicker? I have no idea – but it was interesting.
Los Angeles: 19.1G In | 37.0G OutÂ | 25SÂ – 32S
LA was noteworthyÂ because it was a place with some diversity. Despite the average brew ratio being nearly 1:2, there was the highest percentage of respondents pulling ristrettos here (30%). It also had the biggest variation in brew recipes. LA is home to some diverse espressos, and I think that’s a very good thing indeed – and made me want to go and explore it all a little bit more.
Are Cities Representative of Countries?
No, not really. By and large there was a reasonable (but not huge) difference in the brew styles of a country and its capital city. I’ll use the UK as an example:
|Location||Dose In||Dose Out||Ratio||Brew Time Range|
|UK (excluding London)||18.0g||35.8g||1:2||26sÂ – 32s|
|London||18.5g||34.6g||1:1.87||26sÂ – 35s|
One might think that the trend, the fashion, is for larger espressos. You’d also think that this would manifest first in the capital, but that isn’t the case here. You could argue, however, that the difference between the two is trivial which brings me to my conclusion…
A death of diversity
What shocked me about the replies I got was, ultimately, how similar they all were. I’m really struggling with how I feel about it. Perhaps what we have here is the wisdom of crowds and we’ve actually reached a point of best practice. A 1:2 ratio could very well be the best way to brew an espresso. However, that doesn’t really sit right with me. It’s too neat.
What many of us fall in love with, early in our coffee careers, is diversity. Coffee can taste of so many different things, it feels like it is full of potential and possibility.
I’ve said before that there is a real value to having an aesthetic behind your coffee business. Is what we are seeing here an industry broadly creating an aesthetic behind its espresso brewing? Or are we seeing a trend?
The internet – blogs like this, twitter, newsletters, forums – allow rapid sharing of information. The downside of it is how easy it is that ideas spread very quickly and are often adopted without sufficient testing and rigour.
I’ve been brewing at a 1:2 ratio as a starting point for a while now. This survey has not validated this practice, in fact it has caused me to be deeply suspicious of it. Am I brewing this way because it is mathematically convenient? Am I brewing this way because everyone is? Am I making sure I’m properly exploring recipes and considering what I want to present? I remember all too clearly how much I liked a 1:1.55 ratio…
I believe we’re coming to a point in time where many of us will be reconsidering our approach to espresso. I think it is time to focus and think throughÂ the experience we’re trying to share – from the point of view of the gustatory, the tactile and the intellectual. I don’t believe we all have the same tastes, same ideals, and same goals for our espresso. I don’t believe, even for a second, all consumers want the same thing.
I believe that a homogenous environment is bad for speciality coffee, and bad for the consumer. Going against the grain proudly,Â but mostÂ of all intentionally, gives you a niche in the market. If nothing else, I hope reading this will inspire a little experimentation outside of your current comfort zone. I hope it inspires a conversation with your colleagues to really define clearlyÂ the espresso you most want to source, or to roast, and ultimatelyÂ to serve.