EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
SPECIALITY COFFEE ASSOCIATION OF EUROPE
GOLD CUP RESEARCH
OVERVIEW
In 2010 the Speciality Coffee Association of Europe embarked on a study to ascertain coffee extraction preferences
exhibited by consumers of filter coffee. This was the first public study of its kind since the publication of data in the
United States of America in the 1960’s, under the auspices of the Coffee Brewing Centre. The data from the Coffee
Brewing Centre indicated ideal coffee extraction is when between 18% and 22% of ground coffee is dissolved in hot
water, yielding between 1.15% and 1.35% solubles concentration, or strength, in the final brew.
This specification of correctly brewed coffee has been adopted by various international coffee associations since,
amongst them the Norwegian Coffee Association (NCA) formed in 1975, the Speciality Coffee Association of America
(SCAA) formed in 1982 and the Speciality Coffee Association of Europe (SCAE) formed in 1998.
While the definition of correct strength, being the concentration of dissolved coffee solids in the final brew, differs by
association, all teach the common specification of correct extraction being achieved when between a minimum of 18%
and a maximum of 22% is extracted from the bean and dissolved in the coffee brew. To yield below 18% was deemed an
underdeveloped coffee and to yield more than 22% was deemed to overdevelop the coffee.
The objective of the SCAE study was to establish extraction preference through blind tasting. This was achieved by fixing
the coffee strength of five samples of the same coffee and presenting five different extractions.
ACKNOWLEDGEMENTS
This research, led by the Speciality Coffee Association of Europe Gold Cup Programme would not have been possible
without the kind participation of our industry sponsors, namely:
Primary Sponsors
Secondary Sponsors
1
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
Contents
EXPERIMENT DESIGN 3
RESEARCH LOCATIONS 5
FINDINGS/DUBLIN, IRELAND. FEBRUARY 2011 6
MAASTRICHT, THE NETHERLANDS. MAY 2011 8
COLOGNE, GERMANY. JUNE 2011 10
MILAN, ITALY. OCTOBER 2011 12
COMBINED FINDINGS 14
APPENDICES/CONCLUSION 15
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
EXPERIMENT DESIGN
The SCAE set out to use the existing Coffee Brewing Control Chart as a point of reference. (See Appendix 8 for the CBC
version of the chart and Appendix 9 for the one used by SCAE). The SCAE targeted a fixed strength, being 1.33% across
five different samples of coffee, with each having a different extraction yield, namely 16%, 18%, 20%, 22% and 24%. The
coffee would be brewed using freshly ground coffee and an automatic batch coffee brewer to ensure consistent delivery
at each tasting session.
The coffee and equipment used were to SCAE set guidelines.
90g 85g 80g 75g 70g 65g 60g 55g
1.65%
1.60%
1.55%
STRONG STRONG STRONG
1.50% UNDERDEVELOPED BITTER 50g
1.45%
1.40%
1.35% 45g
UNDERDEVELOPED SCAE IDEAL BITTER
STRENGTH - Solubles Concentration
1.30%
1.25%
1.20%
40g
1.15%
WEAK WEAK WEAK
1.10% UNDERDEVELOPED BITTER
1.05%
14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27%
Extraction - Solubles Yield
Fig A. Five different extractions at consistent strength for sensorial evaluation.
Coffee:
The coffee used had to meet the following criteria:
1. Washed single origin Arabica coffee of speciality grade.
2. Roasted no more than 8 days before the tasting session.
3. The coffee was to be from one roast batch.
4. Roast profile was a medium roast, measured via the Agtron scale of 58 on whole bean and 63 on ground, +/-1 point
(or the equivalent on an alternative scale).
5. The roast was completed in no less than 8 minutes and no more than 12 minutes.
6. The coffee was air-cooled (not water quenched).
3
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
Water:
The brew water used had to meet the following criteria:
1. Hardness – the on site water could fall between 100 and 175 parts per million (ppm).
2. The water used could not vary outside of 10 ppm for the duration of the trial.
3. A taste and odour filter was used.
Grinders:
The grinders used had to meet the following criteria:
1. Deliver a ‘filter grind’, whereby a setting can be achieved to yield at least 45% of the particle distribution
to be between 600 and 800 microns.
2. Be repeatable for at least 10 cycles.
Brewers:
The brewers used had to meet the following criteria:
1. Brew capacity of at least 3 litres.
2. Temperature profile:
a. Deliver a water temperature of between 92°C and 96°C for 90% of the brew cycle.
b. Preferable tolerance of the chosen brew temperature is +/- 1°C.
3. Turbulence – deliver an equal wetting on the coffee bed (this was inspected visually).
4. Deliver a measurable repeatable performance over 10 cycles.
The Set Up:
One grinder was sufficient to run each session. Five brewers were required to run each session.
The inventory for the trials included:
1. At least one grinder.
2. Five brewers to meet the specification set.
3. Roasted coffee.
4. Sample tasting cups.
5. Coded labels to indicate each cup’s extraction.
6. Coded ‘trays’ to vary the presentation of the coffees to the participants.
7. A coffee refractometer to measure the totally dissolved solids in each brew.
8. Participant questionnaires.
Methodology:
1. The five coffees were ‘dialled in’, meaning each brew was checked to ensure they were yielding 1.33%
coffee solids at the five different extractions of 16%, 18%, 20%, 22% and 24%. This was achieved by varying:
a. The dry coffee to fresh water brew ratios.
b. The grind setting on the grinder.
2. Each hour, three litres of each brew was prepared using freshly ground coffee.
3. Each extraction rate was validated using the coffee refractometer.
4. The coffees were held in thermally insulated urns for a maximum of one hour.
5. The cycle was repeated hourly.
6. Each brew was labelled with a non-sequential three digit code, which would be meaningless to the
participant.
7. Each participant would receive a sample of approximately 60ml of each of the five coffees at the same
time, arranged in a random order on the tasting trays (see fig. B).
8. The participant was then asked to score the coffees as per a supplied sheet (see fig. C).
4
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
941 057 418 747 291
TASTE IN THIS ORDER 1
Fig B: SCAE Gold Cup tasting ‘tray’ sample.
Gold cup Research Maastricht June 2011
Age Which of the following do Instructions:
(Circle one as appropriate) you usually add to black • Please taste coffees in the tasting order you have been given.
‹ 25 coffee? • Write the product code in the space given.
25 - 35 (Circle all that apply) • Indicate your likes/dislike for each coffee by ticking one box per now only.
35 - 44 Brown sugar • If you are unsure after reading this, please ask the server for clarification
45 - 55 White sugar before starting.
› 55+ years Artificial sweeteners
Milk Product Dislike Dislike Dislike Like Like Like
Gender Cream Code Extremely Moderately Slightly Slightly Moderately Extremely
(Circle one as appropriate) Soya
Male Non-dairy creamer
Female Other
Nothing
How many cups of coffee
do you drink per day What style of coffee do
(Circle one as appropriate) you usually drink?
(Circle all that apply)
1 - 2 cups per day
3 - 5 cups per day Filter coffee
Instant
› 5+ cups per day
Please insert code
Espresso
Cappuccino
Latte
Stove top
Cafetière/French Press
Fig. C: SCAE Gold Cup participant preference form.
RESEARCH LOCATIONS
The research was conducted using the same protocol in Dublin in Ireland, Cologne in Germany, Maastricht
in The Netherlands and Milan in Italy between February 2011 and October 2011.
The number of samples taken was:
Location Gross
Dublin 188
Maastricht 90
Milan 232
Cologne 131
Total 641
5
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
FINDINGS
Each research location gave us demographic information about the participants along with the preference data.
The preference data received is presented in this section as follows. Scores for each extraction % were weighted.
Extreme like and dislike were multiplied by 8 and -8, moderate like and dislike were multiplied by 2 and -2, and slight
like and dislike by 1 and -1. For each extraction % a sum of the resulting values was calculated and plotted. Bars below
the ordinance indicate more negative preferences (i.e. dislikes).
The findings of the research were as follows:
DUBLIN, IRELAND. FEBRUARY 2011
The Dublin audience demographic:
A predominantly male audience between 25 and 45 years of age, approximately 78% of the Dublin participants
consumed roast and ground coffee daily, with 63.1% of the total consuming filter or French press coffee. The large
majority, 91.1% consume espresso based beverages while a significant majority, 73.8%, add either a sweetener or milk
or a combination to their coffee.
PROFILE Cups per day & accompaniment
5+ CUPS 17.6%
3-5 CUPS 39.4%
1-2 CUPS 42.9%
NOTHING 26.2%
69.4% 30.6% OTHER 0.6%
NON DAIRY CREAMER 0.0%
SOYA 0.0%
CREAM 4.1%
‹ 25 YRS 16.7%
MILK 58.1%
25 - 34 YRS 29.9%
ARTIFICIAL SWEETENERS 1.7%
35 - 44 YRS 26.4%
45 - 55 YRS 20.1% WHITE SUGAR 19.8%
55 + YRS 6.9% BROWN SUGAR 27.9%
COFFEE STYLES
31
0
18.7
36.7
23.4
3.5
22.2
44.4
CAFETIÈRE/
FRENCH BIALETTI
PRESS (ITALIAN
FILTER ESPRESSO CAPPUCCINO LATTE STOVE TOP INSTANT MOKKA) OTHER
Fig. D. Demographic breakdown of Dublin participants.
Note: Coffee styles total more than 100% as many participants consume coffee from a range of preparation methods.
6
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
The majority of Dublin participants preferred a 20% extraction, with 66.1% of participants either liking slightly, liking
moderately or liking extremely this extraction rate. Thereafter the 22% extraction was most preferred.
The extraction with most scores for like extremely was also 20%.
Interestingly, the brew which scored highest in the ‘dislike extremely’ category was the 18% extraction.
20%
PREFERENCES
120
EXTRACTION
100 20%
80
16%
EXTRACTION
60 16% 22%
40
24%
22%
EXTRACTION
20
0
24%
EXTRACTION
-20
18%
Order of Preference
-40
18%
Preference
EXTRACTION
-60
Fig E. Gold Cup preferences Dublin.
7
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
MAASTRICHT, THE NETHERLANDS. MAY 2011
The Maastricht audience demographic:
A predominantly male audience between 25 and 45 years of age, approximately 79% of the Maastricht
participants consumed roast and ground coffee daily. While 58.5% of the total consumed filter or French
press coffee, 93.6% consume espresso based beverages. Again, a significant majority, 74% add either a
sweetener or milk or a combination to their coffee.
PROFILE Cups per day & accompaniment
5+ CUPS 15.6%
3-5 CUPS 42.9%
1-2 CUPS 41.6%
NOTHING 26.0%
76.6% 23.4% OTHER 1.3%
NON DAIRY CREAMER 0.0%
SOYA 0.0%
CREAM 7.8%
‹ 25 YRS 18.2%
MILK 61.0%
25 - 34 YRS 33.8%
ARTIFICIAL SWEETENERS 1.3%
35 - 44 YRS 28.6%
45 - 55 YRS 16.9% WHITE SUGAR 20.8%
55 + YRS 2.6% BROWN SUGAR 22.1%
COFFEE STYLES
35.1
0
32.5
26
19.5
3.9
20.8
39
CAFETIÈRE/
FRENCH
PRESS BIALETTI
(ITALIAN
FILTER ESPRESSO CAPPUCCINO LATTE STOVE TOP INSTANT MOKKA) OTHER
Fig. F. Demographic breakdown of Maastricht participants.
Note: Coffee styles total more than 100% as many participants consume coffee from a range of preparation methods.
8
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
The majority of Maastricht participants preferred a 22% extraction, with 70.2% of participants liking slightly,
liking moderately or liking extremely this extraction rate. Thereafter, the 20% extraction was preferred.
The extraction with most scores for liking extremely was split evenly between 20% & 22%.
Again, the brew which scored highest in the ‘dislike extremely’ category was the 18% extraction.
22%
PREFERENCES
80
EXTRACTION
20%
60
22%
40
EXTRACTION
24%
20
0
20%
EXTRACTION
-20
-40
24% 18%
EXTRACTION
16%
Order of Preference
-60
16% 18%
Preference
EXTRACTION
-80
Fig G. Gold Cup preferences Maastricht.
9
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
COLOGNE, GERMANY. JUNE 2011
The Cologne audience demographic:
A predominantly male audience between 35 and 55 years of age, approximately 97% of the Cologne
participants consumed roast and ground coffee daily. While 47.2% of the total consumed filter or French
press coffee, 79.8% consume espresso based beverages, of which the majority drink espresso. In Cologne,
the majority, 53.8% add neither a sweetener nor milk to their coffee.
PROFILE Cups per day & accompaniment
5+ CUPS 33.7%
3-5 CUPS 48.1%
1-2 CUPS 18.3%
BLACK 53.8%
70.2% 29.8% OTHER 1.0%
NON DAIRY CREAMER 0.0%
SOYA 1.0%
CREAM 3.8%
‹ 25 YRS 8.7%
MILK 45.2%
25 - 34 YRS 20.2%
WHITE SUGAR 12.5%
35 - 44 YRS 29.8%
45 - 55 YRS 32.7% BROWN SUGAR 1.9%
55 + YRS 8.7%
COFFEE STYLES
48.1
7.7
22.1
9.6
2.9
4.8
8.7
38.5
CAFETIÈRE/ BIALETTI
BODUM LATTE (ITALIAN
LOSLICHER
PRESS MACCHIATO MOKKA) OTHER
FILTER ESPRESSO CAPPUCCINO KAFFEE
Fig. H. Demographic breakdown of Cologne participants.
Note: Coffee styles total more than 100% as many participants consume coffee from a range of preparation methods.
10
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
The majority of Cologne participants preferred a 22% extraction, with 62.1% of participants liking slightly,
liking moderately or liking extremely this extraction rate. Thereafter, the 24% extraction was preferred.
The extraction with most scores for liking extremely was 22%.
The brew which scored highest in the ‘dislike extremely’ category was the 16% extraction, with a significant
majority, 70.2%, disliking slightly, disliking moderately or disliking extremely this extraction rate.
22%
PREFERENCES
100
EXTRACTION
22%
24%
50
24%
EXTRACTION
0
-50
18% 18%
EXTRACTION
-100
20%
20%
EXTRACTION
16%
Order of Preference
-150
16%
Preference
EXTRACTION
-200
Fig I. Gold Cup preferences Cologne.
11
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
MILAN, ITALY. OCTOBER 2011
The Milan audience demographic:
A predominantly male audience between 25 and 45 years of age, approximately 96% of the Milan
participants consumed roast and ground coffee daily. While 37.2% of the total consumed filter or French
press coffee, an expected 100% of participants consume espresso based beverages, of which the large
majority, 80.9%, drink espresso. In Milan, a small majority, 51.6% add either a sweetener or milk or a
combination of additives to their coffee.
PROFILE Cups per day & accompaniment
5+ CUPS 27.1%
3-5 CUPS 51.1%
1-2 CUPS 21.8%
NOTHING 48.4%
69.1% 30.9% OTHER 1.6%
NON DAIRY CREAMER 2.1%
SOYA 1.1%
CREAM 2.7%
‹ 25 YRS 8.5%
MILK 17.0%
25 - 34 YRS 30.3%
ARTIFICIAL SWEETENERS 2.7%
35 - 44 YRS 36.2%
45 - 55 YRS 16.0% WHITE SUGAR 21.8%
55 + YRS 9.0% BROWN SUGAR 20.2%
COFFEE STYLES
80.9
0
6.9
2.1
18.6
10.1
3.7
30.3
CAFETIÈRE/
FRENCH BIALETTI
PRESS LATTE (ITALIAN
FILTER ESPRESSO CAPPUCCINO STOVE TOP INSTANT MOKKA) OTHER
Fig. J. Demographic breakdown of Milan participants.
Note: Coffee styles total more than 100% as many participants consume coffee from a range of preparation methods.
12
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
A majority of Milan participants preferred the 24% extraction, with 54.8% of participants liking slightly,
liking moderately or liking extremely this extraction rate. Thereafter, the 20% extraction was preferred.
The extraction with most scores for liking extremely was 24%. This is probably a reflection upon the large
majority, 80.9%, drinking espresso as a general consumption preference – as a 24% extraction is going to
be sensorially the most intense of the brew extractions on offer.
The brew which scored highest in the ‘dislike extremely’ category was the 18% extraction.
24%
PREFERENCES
50
EXTRACTION
40
16%
24%
30
20 EXTRACTION
10
0
16% 20%
EXTRACTION
20% 22%
22%
-10
-20
EXTRACTION
18%
Order of Preference
-30
-40
Preference
18% EXTRACTION
-50
Fig K. Gold Cup preferences Milan.
13
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
COMBINED FINDINGS
On a top level overview, 62% of all participants liked extraction ranges 18-22%, be it slightly, moderately or
extremely. This tends to confirm the Gold Cup standard as being correct. However, one would have expected
a tighter bell curve concentrating more on that 18-22% range based on the conviction with which that
standard is taught worldwide.
The concentration of preferences, 43.8% of participants, lies between 20% and 22%.
OUT- GOLD
SIDE CUP
Fig. L. Gold Cup preferences of the total data set.
Looking at the data as a complete set a little deeper, if we examine the data spread (see fig. M) we will see
the central mark is the median, the edges of the box are the 25th and 75th percentiles and the whiskers are
the most extreme data points.
Here we see 22% proved the most popular. While this is within the Gold Cup range, it is a little surprising
and revealing that this preference is at odds with the niche end of the high grade speciality coffee market,
which currently cites 19%-20% as the correct extraction range.
Gold Cup Research Extraction Preference All Locations
100
50
-50
-100
Preference
-150
16%
1 18% 20% 22% 24%
Extraction %
Fig. M. Gold Cup preferences of the total data set.
14
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
CONCLUSION
In conclusion, the work published in the 1960’s by the Coffee Brewing Centre is both upheld and confirmed
as remaining valid fifty years later. However, current teaching practice and global publications would lead
the market to believe 18-20% are the limits of acceptability, whereas this data would indicate preferences
don’t start and stop at 18% and 22% but rather meet the majority of preferences within this range. We
suspect, faced with a decision to be taken when publishing the original coffee brewing guides in the
1960’s, the Coffee Brewing Centre had to decide amidst data similar to our own where to start and stop the
acceptable range and chose 18% and 22% accordingly.
Paul Stack
SCAE Gold Cup Programme Leader 2008-2012.
June 2013.
APPENDICES
1. Age demographic of participants.
2. Gender demographic of participants.
3. Coffee types consumed.
4. Coffee consumption by volume per participant.
5. Consumption of additives in coffee.
6. Tasting tray.
7. Participant questionnaire.
8. 1965 Coffee Brewing Control Chart by Coffee Brewing Centre.
9. 2007 Coffee Brewing Control Chart by SCAE.
15
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
APPENDIX 1
Age Distribution of Gold Cup Research Participants.
100
80
60
40
25 YRS
20
25 - 34 YRS
35 - 44 YRS
45 - 55 YRS
0 55 + YRS
Maastricht Dublin Milan Cologne
APPENDIX 2
Gender Distribution of Study Participants.
100
90
80
70
60
50
40
30
20
10
Male
0 Female
Maastricht Dublin Milan Cologne
16
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
APPENDIX 3
Preferred Coffee Beverage.
180
160
140
120
100
80
60 Filter
FRENCH PRESS
40 ESPRESSO
CAPPUCCINO
20 LATTE
STOVE TOP
0 INSTANT
Maastricht Dublin Milan Cologne
APPENDIX 4
Cups Per Day.
100
80
60
40
20
1 TO 2
3 TO 5
0 › 5+
Maastricht Dublin Milan Cologne
17
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
APPENDIX 5
Consumption of additives in coffee.
150
100
NOTHING
OTHER
50 NON DAIRY CREAMER
SOYA
CREAM
MILK
ART SWEETENERS
WHITE SUGAR
0 BROWN SUGAR
Maastricht Dublin Milan Cologne
APPENDIX 6
Tasting tray.
941 057 418 747 291
1
TASTE IN THIS ORDER
18
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
APPENDIX 7
Participant questionnaire.
Gold cup Research Maastricht June 2011
Age Which of the following do Instructions:
(Circle one as appropriate) you usually add to black • Please taste coffees in the tasting order you have been given.
‹ 25 coffee? • Write the product code in the space given.
25 - 35 (Circle all that apply) • Indicate your likes/dislike for each coffee by ticking one box per now only.
35 - 44 Brown sugar • If you are unsure after reading this, please ask the server for clarification
45 - 55 White sugar before starting.
› 55+ years Artificial sweeteners
Milk Product Dislike Dislike Dislike Like Like Like
Gender Cream Code Extremely Moderately Slightly Slightly Moderately Extremely
(Circle one as appropriate) Soya
Male Non-dairy creamer
Female Other
Nothing
How many cups of coffee
do you drink per day What style of coffee do
(Circle one as appropriate) you usually drink?
(Circle all that apply)
1 - 2 cups per day
3 - 5 cups per day Filter coffee
Instant
› 5+ cups per day
Please insert code
Espresso
Cappuccino
Latte
Stove top
Cafetière/French Press
APPENDIX 8
1965 Coffee Brewing Control Chart from Coffee Brewing Centre.
19
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
APPENDIX 9
2007 Coffee Brewing Control Chart from SCAE.
90g 85g 80g 75g 70g 65g 60g 55g
1.65%
1.60%
1.55%
STRONG STRONG STRONG
1.50% UNDERDEVELOPED BITTER 50g
1.45%
1.40%
1.35% 45g
1.30%
UNDERDEVELOPED SCAE IDEAL BITTER
1.25%
1.20%
40g
1.15%
WEAK WEAK WEAK
1.10% UNDERDEVELOPED BITTER
1.05%
14% 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27%
20
EUROPEAN EXTRACTION PREFERENCES IN BREWED COFFEE
21
