Professional Documents
Culture Documents
Tns Brand Model
Tns Brand Model
© TNS 2012
Contents
© TNS 2012 2
Why is respondent-level validity important?
The problem of mutually compensating error
Which brands did you buy in
the past three months?
Coca-Cola, Sprite,
and Ariel
The percent who said they used the brand was equal to the percent who actually
used the brand
Yet, many of the people who claimed to use the brand – didn’t
Mutually compensating error: for everyone who says they used the brand but didn’t,
there was another person who said they didn’t use the brand but did
© TNS 2012
The problem is…
... profiling and driver analysis are nonsense!
© TNS 2012 5
This was just one example…
How widespread is this problem?
Not
Awareness First Mention Awareness 0.92 0.57 Valid
Aided Awareness 0.68 0.11
Respondent
Aggregate
Needed
© TNS 2012
How often do people really use the brand they say
they use most?
68
58
36
20
13 8
Don’t Use Use Most Often Don’t Use Use Most Often
Panel Loyalty Card Record
Record
R: Aggregate 0.96
Respondent 0.66
© TNS 2012
Principle (1)
A matching bunch of
0 2931294 9295 8325 5135 3450 2503 1883 1337 1030 818 632 570 396 352 270 217 170 166 97 53 108 1068 2969169
5 12001 2610 1556 870 474 283 150 106 70 40 32 20 24 12 11 5 2 5 2 0 3 14 18290
10 9335 1538 1973 1287 902 557 372 228 168 130 95 66 53 new17 loyals
29 22 13 11 9 3 6 50 16864
15 5542 675 1219 1050 772 555 383 282 176 140 81 81 62 35 37 37 26 23 11 9 6 74 11276
20 3645 380 827 741 662 546 377 307 186 140 113 101 72 60 47 31 23 15 19 7 7 77 8383
25 2595 204 523 546 531 427 368 284 220 154 102 89 61 51 40 33 22 20 14 8 7 68 6367
30 1885 111 312 349 366 346 314 240 181 157 112 80 65 36 44 29 32 15 15 9 10 73 4781
35 1386 80 220 272 275 262 229 226 167 153 101 91 58 45 41 32 25 11 19 7 9 53 3762
40 1068 59 163 172 176 218 189 183 170 125 130 103 68 55 40 33 36 17 14 6 10 64 3099
45 833 38 104 145 172 145 167 132 143 127 104 83 67 76 49 45 32 22 12 8 6 53 2563
50 698 30 79 104 121 107 99 124 134 125 95 83 78 51 50 32 35 20 18 12 7 69 2171
55 560 22 57 72 76 79 92 96 75 89 100 87 79 80 64 41 38 23 17 15 8 55 1825
60 467 17 45 57 67 80 70 67 92 84 79 90 79 77 78 54 47 29 25 16 13 65 1698
65 387 8 40 57 56 64 62 56 72 77 70 79 85 60 55 69 43 30 24 21 13 56 1484
70 26%
297 34%
13 29 slide
42 down
41 63 44 43 57 57 50 61 68 40%
7483 stay
72 Highly
48 41 Loyal
26 14 19 62 1304
75 271 6 24 23 38 38 28 38 46 37 60 42 60 49 66 72 47 42 24 20 11 69 1111
80 196 7 13 20 39 14 26 31 41 45 45 45 46 57 52 58 64 39 41 27 20 73 999
85 170 5 7 14 20 12 30 16 33 23 25 28 37 36 49 In
40 absolute
52 44 terms
41 28 23 62 795
90 118 3 6 13 15 10 7 8 10 14 12 21 28 26 just
31 4% 33 of32 2008’s
32 41 users
35 were
16 71 582
95 60 4 3 6 5 9 12 14 12 9 18 17 18 17 17
loyal in both years
21 18 30 23 37 39 64 453
99 112 0 3 3 10 8 11 6 11 8 9 15 17 14 18 23 16 21 21 22 39 86 473
100 821 5 23 39 51 41 37 39 44 37 42 37 35 54 48 48 51 60 76 41 78 472 2179
Ttl 2973741 15110 15551 11017 8319 6367 4950 3863 3138 2589 2107 1889 1556 1346 1207 1047 872 716 589 398 458 2798 3059628
© TNS 2012
11
The Dynamism of Individual Behaviour
From McKinsey 2002: Financial services and Airlines
Financial Airline
© TNS 2012
So this is the old Conversion Model
A segmentation: Commitment and Availability
18%
15%
12% 11%
6% 4% 8%
61%
43%
27%
13% 13% 16%
9% 6%
© TNS 2012
So this is the old Conversion Model
A segmentation: Commitment and Availability
18%
15%
12% 11%
6% 4% 8%
is incomplete
new users come from
Prediction: Defection and Acquisition
61%
43%
27%
13% 13% 16%
9% 6%
© TNS 2012
Principle (2)
Sales
Consumption
© TNS 2012
Models (2) A General Theory of Brand Relationships
Factors that drive commitment to a brand and the equation
Share
é m æ Share ö ù
Ranki ê å i =1 ç
s
s÷ú
ë è Ranki øû
© TNS 2012
Models (3): Putting the Greek into English
Zipf Equation Step One: Add the scores (10-points, 7-point) and turn into shares and ranks
PMnd Zipf
Examples:
1
Share
R1 (3 brands): 14, 13, 11; Shares: .37, .34, .29 Ranks: 1, 2, 3
é m æ Share
1 öù
Ranki êå i=1 ç
s
s÷ú
R2 (3 brands): 16, 14, 10; Shares: .40, .35, .25 Ranks: 1, 2, 3
ë è Ranki øû R3 (4 brands): 16, 12, 12, 11; Shares: .31, .24, .24, .21 Ranks: 1, 2.5, 2.5, 4
There is still an ‘s’, but it’s set to ‘1’. No exponential transform needed.
Calculate share/rank R1 .37/1 = .37 R2 40/1 = .40 R3 31/1 = .31
.34/2 = .17 35/2 = .18 24/2.5 = .10
.29/3 = .10 25/3 = .08 21/4 = .05
Sum = .64 Sum = .66 Sum = .62
[.37; Share
.34; .29] [.40; Share
.35; .25] [.31; Share
.24; .21]
é m æ Share ö ù 58; 27; 16 é m æ Share ö ù 61; 27; 12 é m æ Share ö ù 60; 15; 15; 10
êå i=1
[1; 2; 3] .64s ÷ ú
Xçè Rank å i=1Xçè Rank
[1; 2; ê3] .66s÷ú êå i=1
[1; 2.5; 4] .62s ÷ø ú
Xçè Rank
ë i øû ë i øû ë i û
© TNS 2012
All of this is well known…
… award winners
Ipsos ‘Wallet
Allocation Rule’
Rank 2
Attitudinal Equity (1- )´( )
1
No Brands+1 No Brands
é m æ 1 öù
Rankis ê å i=1 ç ú
ë è Rankis ÷ø û
© TNS 2012
How the new Conversion Model compares
… a more than 20% improvement
1
Ipsos ‘Attitudinal
R = 0.59 é m æ 1 öù
Equity’ Rankis ê å i=1 ç s÷ú
ë è Ranki øû
Share
New ‘Power in the
R = 0.62 é m æ Share ö ù
Mind’* Rankis ê å i=1 ç ú
ë è Rankis ÷ø û
© TNS 2012
The Survey: Questions that measure Commitment and predict
the share of wallet that a person will give each brand
© TNS 2012
Models (4): How to quantify Power in the Market
Market Drivers: Push (presence), Barrier (weakness, failure)
© TNS 2012
Models (5) Driver Analysis
How many Touch-points or Attributes do you need
No, no won't do. Only picture to yourself a nice soft wife on a sofa with good
fire, & books & music perhaps. Marry, marry, marry QED
The makers of this brand understand my needs in a laundry detergent product Delights my senses (sight, smell, feel) Makes it easy to figure out what product is right for me
This brand does what it promises Offers scents that I wish were available in other products besides laundry detergent I feel a connection to other people who use this brand
This is a brand I can trust Has scents that are better than any other brand This brand has a clear and distinct point of view
This brand provides the right level of cleaning for me Offers the best variety of scents This brand aims to fulfill a purpose beyond mere profit making
Is an excellent product for the money Leaves laundry with a long-lasting scent I can identify with the values that this brand represents
Simplifies the laundry process Provides scents that make doing laundry more enjoyable The values that this brand stands for are highly relevant to me
Is a brand I would recommend my friends to buy Has a scent I prefer This brand inspires me to talk about the brand and its values
Keeps colors bright Offers scents I canÆt get enough of I engage with others about the values that this brand represents
Leaves clothes smelling fresh Has great-looking packaging Removes stains better than any other detergent
Keeps clothes looking their best The best smelling way to get laundry done Cleans better than any other detergent
Is good for all the laundry I do Makes the laundry experience more enjoyable than I would expect Is an authority on taking care of fabrics
Is a reliable laundry detergent Helps me feel in control of the laundry process Prevents colors from fading better than any other detergent
Is an effective laundry detergent Helps me feel I am taking good care of my family Provides superior whiteness
Provides a deep down clean This brand makes me feel confident This brand makes my clothes soft
Provides superior results in any type of load I feel like this brand acts in my best interest Keeps clothes looking like new
Is effective at removing odors This brand makes me feel proud of my laundry results Provides benefits that justify the price
Leaves clothes feeling their best This brand helps me feel good about myself I love this brand
Gets my clothes clean even in cold water Allows me to present my family at their best I would miss this brand if it went away
Works on clothes of all colors Helps me to always make a good impression with my appearance I try to buy this brand whenever I can
Is effective at removing stains Makes fabrics more enjoyable to wear This is pretty much the only brand I would buy
I always have a consistently positive experience with this brand Provides thoughtful solutions to my own needs Helps me get my familyÆs recognition
Everything I see from this brand is consistent (in store, TV, online, etc) This brand is innovative Is a pleasure to shop for
This brand provides a more natural solution for detergent Is a good member of the community Is youthful
Helps me feel I am spending my money wisely by buying this brand This is an honest brand Is safe for those with sensitive skin
This brand makes me feel like a smart shopper Is currently a leading brand Does not cause allergic reactions
Brands
Personally relevant brands (3 – 7)
Attribute1 …
Attribute2 …
Attribute3 …
Attribute4 …
… … … … … … …
Attributej …
© TNS 2012
Models (5a):
(2): Simple
Simpletransformations
transformationsthat
thatimprove
improvevalidity
validity
Weighted share
Unweighted share
of of
mentions…
mentions…
Attribute1
100 …
Attribute2 50
50
…
Attribute3
33
33
33 …
Attribute4
50
50 …
… … … … … … …
Attributej 100 …
© TNS 2012
Models (5b):
(2): Simple
(2a): Simpletransformations
transformationsthat
thatimprove
improvevalidity
validity
Unweighted
Weighted share
share
of of
mentions…
mentions…
Attribute1
100 …
Attribute2 50
57
50
43
…
Attribute3
33
50
33
38
33
12 …
Attribute4
50
20
50
80 …
… … … … … … …
Attributej 100 …
© TNS 2012
There is one more crime to which I must refer..
The obvious idiocy of straight-line modelling
P = 0.64
P = 0.28 Day-to-day
P = 0.08 Use
Relevant
Endorsement
Market
Visibility
Planned
Power in the Mind Communications
Commitment
Relevant Brands
Consider 54s Salience
Stated Share
Commitment to All
Performance 46s
Relevant Advantage
Involvement Personal Connection
Brand Drivers Image
Relevant Attributes 120s
Share of Mentions Motivation
Market Drivers
Push Factors 48s Presence
Market Barriers
© TNS 2012
Let’s summarize…
Key principle (1): Stop asking questions that don’t predict very well
Key principle (2): Stop torturing respondents – ask them what’s relevant
Key principle (4): Tie everything to market share and market share flows
New markets
4
3
More customers
2
New New products
customers & services
1
Today’s Loyalty &
business new spend
3.9%
2%
Current market share
Opportunity
5.4%
7%
Implicit market share
Increase spend from
existing users
Decreased spend by
1.5%
Gains are most likely to
existing users At risk come from new users
-0.3% -0.5%
Defection
defections Steady
-0.2% 2.0%
Decrease Increase
Waitrose
1.2% 1.2%
+4.9%
0.8%
0.6%
… and how much each one is worth: net GBP 5.1 billion
Waitrose
1.5%
1.3%
+4.9%
0.6%
0.3% 0.3%
Your Your
Business Competitors
Attributes
Waitrose Tesco Asda Sainsbury’s Morrison's