Evaluating whole cup experience in gourmet espresso coffee by

Evaluating whole cup experience in gourmet espresso
coffee by using dynamic methods
DUGAS Valentine*, PINEAU Nicolas**, FOLMER Britta*
*Nestlé Nespresso, Lausanne, Switzerland; **Nestlé Reasearch Center, Lausanne, Switzerland
INTRODUCTION
Highest quality coffee beans meet perfect physical quality and specific taste profiles which are obtained through interplay of plant variety, soil and climate. The art of
blending and roasting of such quality coffee beans allows the development of unique, rich and complex sensory profiles.
Traditionally, in-cup coffee description is obtained using monadic profiling. However, this methodology does not give a dynamic view of the whole cup experience.
Temporal Dominance of Sensations (TDS) has already been used in this context to better understand the consumer sensory journey through the whole product (1-2).
In the current study, we have applied a modified TDS methodology aiming at differentiating 12 gourmet espresso coffees beyond the classical Quantitative Descriptive
Analysis (QDA) tasting, and providing new information on the dynamics of coffee tasting over time.
MATERIAL AND METHOD
Method: Capture the dominant attribute for each sip. 1 sip / 30 sec.
• 14 trained panelists
• 12 commercially available Nespresso coffees (3 blocs of 4 randomized products)
• 2 repetitions
Adapted TDS methodology to fit to the evaluation of a whole cup of espresso (3):
• 40ml espresso extracted from a Nespresso machine
• 7 sips in total, consuming 1 sip every 30 seconds
• Choice of the dominant attribute among a list of 10 pre-selected attributes of texture, aroma and taste (4)
(dominant attribute defined as “the one triggering the most attention” (5)).
TDS ADDS A DIMENSION TO THE COFFEE TASTING EXPERIENCE
E1
40
30
20
20
10
10
0
0
1
2
3
4
5
6
1
7
E3
40
2
3
30
30
20
20
10
4
5
6
7
5
6
7
5
6
7
E4
40
10
0
0
1
2
3
4
5
6
7
1
E5
40
2
3
30
30
20
20
10
4
E6
40
3’
4’
5’
Extraction
and service
Start
extraction
1st sip 2nd sip 3rd sip 4th sip 5th sip 6th sip 7th sip
Schematic illustration of the method and timing of the evaluation
Different TDS profiles are obtained across products, allowing the products to be
differentiated in a different dimension as compared to QDA.
Looking at the sequence of attributes along consumption, we observe some trends:
• Roasted notes are triggering the attention in the first sip only.
• Bitterness only appears towards the end of the cup.
• Acidic coffees in general have a maximum of 3 dominant attributes. Balanced and
more bitter coffees have more attributes varying along the cup.
• Textural attributes never triggered the attention, even if in QDA profiling they appear
as differentiating aspects.
• More specific attributes such as cereal, fruity, acid appear and vary all along the
consumption.
E2
40
30
2’
0
10
0
0
1
2
3
4
5
6
7
1
E7
40
2
3
40
30
30
20
20
10
10
0
4
E8
TDS HELPS DIFFERENTIATING COFFEES
RISTRETTO NC 25ml
0
1
2
3
4
5
6
7
1
E9
40
2
3
30
20
20
10
10
0
5
6
E10
7
T-Silky, Smooth
0
1
2
3
4
5
6
7
30
20
20
10
10
0
0
1
2
3
4
2
3
5
6
5
6
Roasted
7
Fine Acidity
1
7
4
E12
40
30
Bitter
O,F-Sweetish
1
E11
40
E10
T-Body
E10
40
30
4
2
3
4
5
6
O,F-Carbony
O,F-Fruity / Floral
7
E11
O,F-Dry Vegetal
O,F-Green Vegetal
Acid
Bitter
Cacao
Cereal
Fruity/Floral
Gritty
Roasted
Smooth
Thick
Woody
0
Curves obtained by summing -up significant dominance scores all along consumption.
X-axis = sips ; Y-axis = dominance rate (cumulative)
ROMA NC 25ml
E11
T-Body
T-Silky, Smooth
Bitter
Sensory profiles obtained through adapted TDS
X-axis = sips ; Y-axis = dominance rate (cumulative)
FROM HIGHLY COMPLEX TO VERY SELECTIVE AROMATIC NOTES
COSI NC 40ml
ARPEGGIO NC 25ml
T-Body
T-Body
T-Silky, Smooth
O,F-Sweetish
E1
T-Silky, Smooth
Bitter
O,F-Sweetish
Roasted
Fine Acidity
O,F-Carbony
O,F-Fruity / Floral
O,F-Dry Vegetal
O,F-Green Vegetal
E8
O,F-Sweetish
Roasted
Bitter
Roasted
Fine Acidity
Fine Acidity
O,F-Carbony
O,F-Carbony
Cocoa
O,F-Fruity / Floral
O,F-Dry Vegetal
O,F-Fruity / Floral
O,F-Dry Vegetal
O,F-Green Vegetal
O,F-Green Vegetal
Sensory profiles obtained
through classic QDA
Although standard sensory profiling would imply that
E10 and E11 are very close, the TDS methodology
shows that they present clearly different sequences
of dominance along consumption. Fruitiness, acidity
and cereal notes did not appear in the same order
and bitterness is more triggering in the last sip of
E10 than in E11.
CONCLUSION
TDS curves and classic profiles of E1 and E8
The current study shows not only that dominant attributes vary along the consumption
of one cup, but also that for some coffees it was difficult to find a consensus on one
dominant attribute for each sip.
The graph shows the example of 2 very different coffees. Very few attributes reach the
significance threshold for E8 but 9 out of the 10 attributes are selected at least by one
panelist. In E1, there are 3 very significant triggering attributes and the others are
almost not mentioned.
The current study shows that an adapted TDS methodology can give additional
information in comparison to classic QDA. Looking at the sequence of attributes that
trigger the attention, the differentiation between products is higher. Moreover, while some
coffees have very few and selective attributes along most of the cup, others vary
continuously throughout the consumption.
It could be hypothesized that a high number of significant dominant attributes could be an
indicator of a complex coffee, while few dominating attribute would indicate a coffee with
pure and very specific sensory aromatic notes.
REFERENCES:
(1) Lenfant et al., Perception of oral food breakdown. The concept of sensory trajectory, Appetite 52, 2009, 659–667
(2) Vandeputte et al, Innovative methods to assess the evolution of the sensory characteristics during the tasting of a full product portion, poster in 9th Pangborn Sensory Symposium
(3) Baron et al., Impact of crema on the aroma release and the in-mouth sensory perception of espresso coffee, Food & Function, 2012, 3 (9), 923 – 930
(4) Pineau et al., Temporal Dominance of Sensations: What is a good attribute list? Food Quality and Preference 26 (2012) 159–165
(5) Saint-Eve et al., Impact of panel training, attributes list, type of response and dominance definition on TDS response, poster in 9th Pangborn Sensory Symposium