Sensory Analysis of Nondairy Frozen Desserts: Flavor still Rules
MaryAnne DrakeNorth Carolina State University
Southeast Dairy Foods Research Center
The Bottom Line
• Competitive and expanding market
• What are the keys to success?
Keys to success: Exploiting flavor advantages
Know your customer(s):– Different customers have different desires
Know your ingredient(s)– Understand functionality– Understand ingredient flavor and flavor stability– Understand how dairy and nondairy sources are distinct
Know your applications– Different applications have different flavor carry-through potential
Understanding the customer
• What do consumers want?
• Consumers are looking for variety and good-for-you indulgence
• What are the flavor issues?
Consumer Disconnect
Bored
Calm
DisappointedEager
Energetic
Free
Friendly
Glad
Good
GoodNatured
Guilty
Happy
Joyful
Loving
Merry
MisledNostalgic Peaceful
Pleasant
Pleased
SatisfiedUnsatisfied
Warm
Whole
Blue Bell B
Blue Bell UB
Great Value B
Great Value UB
Haagen Daz B
Haagen Daz UB
Halo Top B
Halo Top UB
Party Pail BParty Pail UB
-0.5
-0.3
-0.1
0.1
0.3
0.5
-0.5 -0.3 -0.1 0.1 0.3 0.5 0.7
F2 (2
1 %
)
F1 (39 %)
Attribut es Pr odu ct s
Consumer emotions for vanilla desserts: Unbranded and Branded CATA
N=100 consumers
Consumer Disconnect
Bored
Calm
DisappointedEager
Energetic
Free
Friendly
Glad
Good
GoodNatured
Guilty
Happy
Joyful
Loving
Merry
MisledNostalgic Peaceful
Pleasant
Pleased
SatisfiedUnsatisfied
Warm
Whole
Blue Bell B
Blue Bell UB
Great Value B
Great Value UB
Haagen Daz B
Haagen Daz UB
Halo Top B
Halo Top UB
Party Pail BParty Pail UB
-0.5
-0.3
-0.1
0.1
0.3
0.5
-0.5 -0.3 -0.1 0.1 0.3 0.5 0.7
F2 (2
1 %
)
F1 (39 %)
Attribut es Pr odu ct s
DISLIKE
Consumer emotions for vanilla desserts: Unbranded and Branded CATA
N=100 consumers
Frozen Dessert Flavor Challenges
• Flavor differences with different bases• Flavor and texture differences with fat removal• Flavor and texture differences with sweeteners• Flavor changes as you increase protein content • Flavors due to other ingredients (solids, stabilizers)
• Cereal• Sulfurous• Nutty• Beany• Pyrazine/ bell
pepper• Earthy/soil• Fruity• Sour aromatic• Sour taste• Umami taste
• Sweet aromatic• Milky/ cooked• Milkfat
• Cardboard• Brothy• Fatty• Grassy• Salty taste• Bitter taste• Astringency
MilkPlant liquids
Flavor Differences between Plant Beverages and Milk
Flavor profile of milk/cream-based mix for frozen desserts
0.0
1.0
2.0
3.0
4.0
Cooked/milky
Beany
Cardboard
Cereal
Roasted nutty
FattySalty taste
Sweet taste
Bitter taste
Umami taste
Astringency
Milk
Flavor profiles of milk and plant-based mixes for frozen desserts
0.0
1.0
2.0
3.0
4.0
Cooked/milky
Beany
Cardboard
Cereal
Roasted nutty
FattySalty taste
Sweet taste
Bitter taste
Umami taste
Astringency
MilkLegume beverageAlmond beverageSoy beverage
Flavor profiles of vanilla ice cream
0.0
1.0
2.0
3.0
4.0
5.0
vanilla
Cooked/milky
Coconut
Beany
Cardboard
Nutty
Salty taste
Bitter taste
Metallic
Astringency
Premium DairyConventional Dairy
Flavor profiles of vanilla ice cream and nondairy frozen desserts
0.0
1.0
2.0
3.0
4.0
5.0
vanilla
Cooked/milky
Coconut
Beany
Cardboard
Nutty
Salty taste
Bitter taste
Metallic
Astringency
Premium DairyConventional DairyAlmond baseCoconut base
Temporal flavor profiles of vanilla ice cream versus vanilla coconut frozen dessert
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100
Cita
tion
Prop
ortio
n
Time (s)
Bitter
Cardboard
Chalky
Cloying Mouthcoating
Coconut
Creamy/Smooth
Metallic
Milky/Cooked
Salty
Vanilla
Some typical differences:
Milky
Coconut
BitterCloyingmetallic
Other coconut specific attributes
Temporal flavor profiles of vanilla ice cream versus vanilla almond/legume frozen dessert
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100
Cita
tion
Prop
ortio
n
Time (s)
Bitter
Cardboard
Chalky
CloyingMouthcoatingNutty/Beany
Creamy/Smooth
Metallic
Milky/Cooked
Sweet
Vanilla
Milky
Nutty/beany
MetallicCloyingBitter
Frozen Dessert Challenges • Flavor differences with different bases
– Other flavors present with plant bases– Vanilla flavor is the most challenging– Texture differences
More Frozen Dairy Dessert Challenges • Flavor differences as you decrease fat and sugar
– Fat plays a critical role in flavor release– Alternate sweeteners have distinct temporalities and
other sensory attributes (metallic and bitter)– Increasing protein increases astringency
• Texture issues– Melting rate– Solids (and other solids may bind flavorings and
contribute other flavors)
0.0
1.0
2.0
3.0
4.0
Vanilla
Cooked/milky
Cardboard
Salty tasteBitter taste
Metallic
Astringency
All Natural Control
Flavor profiles of vanilla ice cream and frozen dairy desserts: Decreasing fat and sugar
0.0
1.0
2.0
3.0
4.0
Vanilla
Cooked/milky
Cardboard
Salty tasteBitter taste
Metallic
Astringency
All Natural Control
Lowcarb 1
Lowcarb 2
Flavor profiles of vanilla ice cream and frozen dairy desserts: Decreasing fat and sugar
0.0
1.0
2.0
3.0
4.0
5.0
Vanilla
Cooked/milky
Cardboard
Salty tasteBitter taste
Metallic
Astringency
All Natural Control
Lowcarb 1
Lowcarb 2
Reduced fat and sugar 1
Reduced fat and sugar 2
Reduced fat and sugar 3
Flavor profiles of vanilla ice cream and frozen dairy desserts: Decreasing fat and sugar
Impact of decreased sugar and fat
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 20 40 60 80 100
Cita
tion
Prop
ortio
n
Time (s)
All NaturalLow Carb 1Low Carb 2Lactose FreeReduced Fat and Sugar 1Reduced Fat and Sugar 2Significance Limit
Temporal perception of cardboard flavor in vanilla frozen dairy desserts
All natural vanilla Ice cream: no cardboardflavor
Impact of decreased sugar and fat
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 20 40 60 80 100
Cita
tion
Prop
ortio
n
Time (s)
All NaturalLow Carb 1Low Carb 2Lactose FreeReduced Fat and Sugar 1Reduced Fat and Sugar 2Significance Limit
Temporal perception of bitter taste in frozen dairy desserts
Reduced fat and sugar 1
Impact of increasing protein
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 20 40 60 80 100
Cita
tion
Prop
ortio
n
Time (s)
All NaturalLow Carb 1Low Carb 2Lactose FreeReduced Fat and Sugar 1Reduced Fat and Sugar 2Significance Limit
Temporal perception of astringency
Astringency perception increasesAcross the eating experience
Pronounced with higherProtein and non nutritivesweeteners
The good news and the challenges• Consumers are seeking healthy options – in all food
sectors• Different challenges exist with different ingredients and/or
ingredient substitutions• Flavor and functional challenges
– Formulation challenges: these products must deliver desirable flavors and textures
FLAVOR STILL RULES
Summary
• Significant flavor and texture challenges exist with formulating alternative frozen desserts
• An understanding of the challenges specific to the application is necessary
• Careful selection of ingredients and formulation
Moving forward• Understand your consumer and your competition:
SENSORY ANALYSIS• Carefully consider and screen protein sources
– Tremendous flavor variability• Careful selection of stabilizer systems and flavors
– Vanilla is the ultimate challenge– Masking agents may be useful
• Flavor and texture are sensory perceptions
• Instruments do not document flavor and texture
Sensory analysis is a requirement
Where is Sensory Used?
MarketingQuality AssuranceResearch & Development
INT
RO
DU
CT
ION
TO
SE
NS
OR
Y
Flavor
• Flavor - perceptions perceived via the chemical senses from a product in the mouth. Flavor includes:– Aromatics– Basic Tastes
• sweet, salty, sour, bitter, umami– Trigeminal stimulation (chemesthesis)
Source: http://www.enologyinternational.com/articles/senses.html
Aroma/Flavor Physiology
umami
Trigeminal nerves(5th cranial nerves)
Things to consider• Why do sensory tests have to be conducted a certain way
and with certain scales?
– Human physiology – how we perceive external stimuli
– Psychometrics – how humans respond to external stimuli
• For actionable results, the correct test should be conducted in the correct manner
Gustation (Basic Tastes)• Along with olfaction, the most widely studied chemical
sense• Perceived by taste buds • Taste buds are balls of modified epithelial cells with a
pore or channel. – Solutions make contact with cells through the pore– Lifespan of 1 week
• Taste buds are located within the papilla on tongue and soft palate
• Different tastes are not perceived on different areas of the tongue
The Sense of Taste• Taste: the ability to respond to dissolved molecules and
ions• Physiology
– Humans detect taste with taste receptor cells. These are clustered in taste buds
– A single taste bud contains 50-100 taste cells representing all 5 taste sensations
Physiology of Olfaction• Stimulation of the olfactory epithelium by a volatile compound• Comprises the majority of what we consider flavor• There are ca. 1000 nerve bundles in the olfactory bulb• Only a fraction of what we sniff or chew is actually “sampled”• Receptors are true nerve cells rather than modified epithelial
cells• More rapid response than basic tastes – also more rapid
adaptation
UN
DE
RS
TA
ND
IN
G T
HE
SE
NS
ES
Olfaction• Optimum sensitivity in 1-2 seconds• Adaptation (decreased sensitivity) rapidly occurs
– Issue for threshold tests, odor recognition, or odor profiling
• Enormous discriminating power• Less sensitive to discrimination of intensities
Chemesthesis• Stimulation of the trigeminal (cranial) nerves in the
mucous membranes of the nose, eyes, mouth, and throat
• Plays a critical role in acceptability of certain foods and pharmaceutical products
Practical Application: Physiology• Sufficient tasting volumes• Lids on containers • Appropriate temperatures• Appropriate procedures for orthonasal evaluation• Warm up/calibration samples
Sensory RulesBe on Time
No Smoking, Eating, Drinking• 15-30 minutes before panel
Avoid Distractions
Don’t wear cologne, perfume,
etc.
Wash your hands before tasting
Pay attention & participate
IN
TR
OD
UC
TI
ON
T
O
SE
NS
OR
Y
Sensory Analysis Tools• Quality judging• Analytical tests• Affective tests• Lots of different types of tests – which is the right one
to use?– Clear objective(s) are crucial– Hammer vs screwdriver
Sensory Analysis: a vast tool box
• Analytical tools: objective, use trained panelists, require replication, descriptive analysis is the most powerful tool in this box
• Consumer tools: subjective, generally require large numbers of consumers
• Right tool(s) for the right objective can generate powerful actionable results
Analytical Sensory Tests
• What’s their purpose?– Identify defects or deviations– measure or evaluate product attributes– Focus is on the product whereas consumers tests
focus on the consumer– Because the focus is on the product, these tests
use screened or trained panelists– And liking is NOT an issue
• Sensory tests that are:– analytical in nature– measure quantifiable data– panelists are screened and/or trained
• Examples:– difference (discrimination) tests– Threshold tests– descriptive sensory analysis
Analytical Sensory Tests
• Sensory tests that:– measure or evaluate consumer perceptions and/or
responses– Do not use trained panelists!– Are not replicated (by panelists)– Generally require large numbers (>50) of consumers– Includes quantitative and qualitative tests
Affective (Consumer) Sensory Tests
Consumer tests• Quantitative
– Acceptance/preference– Surveys
• Qualitative– Focus groups– Interviews/MEC
• Acceptance tests– Panelists asked to score acceptance of a product– Preference can be inferred – Multiple attributes can be assessed– Quantitative measure of degree of liking
Quantitative Consumer Tests
• Acceptance tests– Panelists asked to score acceptance of a product– Preference can be inferred – Multiple attributes can be assessed– Quantitative measure of degree of liking
Quantitative Consumer Tests
Acceptance tests
• You can ask more than liking:• Purchase intent• Intensity questions
– Correlations with liking attributes• JAR (just about right) questions
– Penalty analysis• Comments (likes and dislikes)• Just remember: consumers are best at telling you what they like and
don’t like
Acceptance tests
• Ballot design is crucial– What you ask and how you ask it– Example: off flavors
• Overall liking should always be asked first• Randomized and/or balanced order of presentation
Surveys
• Acceptance tests can be pricey to obtain high N
• Online surveys can provide alternatives
Surveys• Can be used for:
– Product concepts– Brands– Labels– Emotions– Purchase habits
Surveys• Same principles apply: screen for appropriate target
consumers• What you ask and how you ask it is crucial• Lots of question format options*
– *many of these are also appropriate for acceptance ballots/screeners also
Surveys• Surveys can look like traditional acceptance test ballots
but there are other design options– Conjoint analysis: evaluate combinations of concept
attributes/label claims– MaxDiff analysis: evaluate single concept
attributes/label claims– TURF – identify maximum reach and frequency for
concepts
Drivers of liking• Also called preference mapping• Modeling or predicting consumer liking responses• Can be done with:
– Consumer scores alone – internal preference mapping– Consumer scores and additional (external) data, usually descriptive
profiles – external preference mapping– Done using multivariate modeling approaches including PLS, MLR,
and LSA– Consumer responses may or may not be clustered prior to modeling
liking
Qualitative Consumer Tests• Uncovering and characterizing consumer needs, desires,
and perceptions– What drives initial purchase and repeat purchase for a
particular product• How do we do this?
– Talking, listening, behavioral observations– Not quantitative – can’t scale this
• Interviews– MEC (means end chain)
• Focus groups• Focus panels, modified groups
– Napping• Hybrid techniques
– QMA
Qualitative Tests
Acknowledgements
Sensory Service Center