Transcript Assessing Ad Message Effectiveness

Assessing Ad
Message
Effectiveness
“Watchability” by Millward
Brown
• Function of ‘involvement’ and ‘enjoyment’
• Use
–
–
–
–
–
–
Humor
Music
Limited voice-overs
Pace
Celebrities
Cute things
• More likely to be recalled and drive sales
2
Overview of Advertising Research
It’s not easy or inexpensive
but the value outweighs the drawbacks.
• More than 80% of advertisers and agencies
pretest television commercials before airing
them on a national basis.
3
What Does Advertising Research
Involve?
Media Effectiveness (covered later)
Message Research
• To test effectiveness of messages
• Pretesting ads during developmental stages
• Posttesting to determine if messages achieve
their established objectives
4
Four Stages At Which Ad Message
Research Might Be Conducted
1. Copy development stage (testing ideas)
2. “Rough” stage (testing rough concepts
and layouts)
3. Final production stage (testing complete
ads)
4. After the ad has been run in the media
(post testing)
5
Industry Standards for Message Research
PACT Principles
Principle 1: provide measurements that are relevant to
the advertising objectives
What question would you use to test recall?
Principle 2: requires agreement about how the results
will be used in advance of each specific test
e.g. specify what the results of the focus group will be
used for.
Principle 3: provides multiple measurements because
single measurements are generally inadequate
Think of two ways to test brand preferences
6
Industry Standards for Message Research
PACT Principles
Principle 4: based on a model of human response to
communications
Reception of a stimulus > Comprehension of the
stimulus > The response of the stimulus
Check whether consumers understood the ad before
forming conclusions based on sales
Principle 5: allows for consideration of whether the
advertising stimulus should be exposed more than
once
external generalizability
7
Industry Standards for Message Research
PACT Principles
Principle 6: recognizes that a more finished piece of
copy can be evaluated more soundly
Alternative executions be tested in the same degree of
finish
Principle 7: system provides controls to avoid the
bias normally found in the exposure context
Increase internal validity of results
8
Industry Standards for Message Research
PACT Principles
Principle 8: takes into account basic considerations
of sample definition
Requires that the sample be representative of the target
audience
Principle 9: can demonstrate reliability and validity
A reliable test is one that yields consistent results
A valid test is one that is predictive of the marketplace
performance
9
What Can Be Learned from
Message Research?
• Message research is needed to diagnose
an advertisement’s prospective equityenhancing and sales-expanding potential.
• The ARF assessed which of 35 measures
best predicts the sales effectiveness of
television commercials. The only definitive
conclusion of their study is that no one
measure is universally appropriate or best.
10
Message Research Methods
Qualitative Message
Research
Quantitative
Message Research
11
Qualitative Message Research
•
•
•
•
•
Focus Groups
In-depth Interviews
Observation
Ethnography
Projective Techniques
– ZMET
• Neurological Networks
12
Illustrative Message Research
Methods
13
Advertising Research Methods
Message
Research
Methods
Recognition & Recall
Physiological Arousal
Persuasion
Sales Response
14
Message Research Methods
• Starch readership
service (magazines)
• Bruzzone tests (TV)
• Burke day-after recall
(TV)
Recognition & Recall
Physiological Arousal
Persuasion
Sales Response
15
Starch Readership Service
•Examines reader awareness of ads in consumer magazines
and business publications
•Measures the primary objective of a magazine ad—to be seen
and read
•Over 75,000 ads in more than 140 publications surveying
more than 100,000 people annually. Sample sizes – 100 to 150
nos. per issue
•Interviews in the early life of the publication
•Respondents asked if they read an ad and what they
remember from the ad
•Readers are classified
16
Starch Readership Service
Classifications
Noted—percentage who remember
having previously seen the ad in the issue
being studied
Associated—percentage who noted the
ad and saw or read some part of it that
clearly indicated the name of the brand or
advertiser
17
Starch Readership Service
Classifications
Read Some—percentage who read any
part of the ad’s copy
Read Most—percentage who read half
or more of the written material in the ad
18
Starch Readership Service
Indices are developed for the four classifications.
Two types of indices:
1. Ad’s scores against average of all ads in the magazine issue
2. ADNORM index: compares an advertisement’s scores
against other ads in the same product category as well as the
same size (e.g., full page) and color classifications (e.g., fourcolor ads)
Average Adnorm index is 100. A score of 130 indicates the ad
did 30% better than the comparable ads.
19
Measures of
Recognition &
Recall
• 39% of people
noted the ad,
• 37% associated it,
• 27% read some of
it, and
• 10% read most of
the body copy
contained in the
ad.
20
Measures of
Recognition &
Recall
• 58% of people noted
the ad,
• 54% associated it,
• 39% read some of it,
and
• 39% read most of
the body copy
contained in the ad.
21
Starch Readership Service
• Kia Sorento ad
Adnorm index scores
• 72 (noted)
• 74 (associated)
• 73 (read some)
• 59 (read most)
• The ad performed
28% worse than the
average noted score
for similar ads.
• Jose Cuervo Especial
Tequila ad
• 107 (noted)
• 108 (associated)
• 105 (read some)
• 229 (read most)
• The ad performed
slightly above
average noted score
for similar ads.
22
Problems with the Starch scores
• Respondents may report ads that actually
never ran or may have run in another
magazine / issue
• Respondents may report that they have
not seen an ad when in fact they have.
– Implicit memory
– Effect on attitudes in spite of lack of conscious
recollection
23
Bruzzone Tests
Bruzzone test provides
advertisers with a test of
consumer recognition of
television commercials.
Anything identifying the
brand is removed, to test
whether viewers
remember the name of the
advertiser.
24
Bruzzone Tests
• Emails 15 commercials (six frames from the commercial) to a
sample
• Each respondent is asked if they remember seeing the commercial
on TV
• If yes – more questions; If no or not sure – skip to next.
• Brand name references removed from the frames
• Also asked how the commercial made them feel (check from a
battery of adjectives); how interested they were in it, whether they
liked it, etc.
• Finally asked if they remember the brand name
• http://www.bruzzone-research.com/ONLINE_DEMOq.htm#Q26
25
Measures of Recognition & Recall
Bruzzone test
Advertising Response Model (ARM) links
responses to the 27 descriptive
adjectives to consumers’ attitudes toward
both the ad and the advertised brand and
to purchase interest.
Budweiser “Thanking the Troops”
http://www.youtube.com/watch?v=LawSGh
TM0es
26
ARM for the “Thanking the Troops”
Commercial
27
Taco Bell Carne Asada Superbowl commercial
http://www.youtube.com/watch?v=7oN7X9z6Cc
4
4
28
Advertising Response Model (ARM) for the
“Carne Asada Taquitos” Commercial
29
Measures of Recognition & Recall
Bruzzone test
Provides valid prediction of actual marketplace
performance along with being relatively
inexpensive
Doesn’t provide a before-the-fact indication
Tests offer important info for evaluating a
commercial’s effectiveness and whether it
should continue to run
Tests are performed online
30
Burke Day-after Recall
• Used to assess the effectiveness of test
commercials
• Tested the day following the day the
commercial is first aired
• Telephone interviews
• Product / Brand cue given
• Asked if they saw the commercial and
then asked to recall all they can about it.
31
Measures of Recognition & Recall
Burke Day-After Recall Testing
(1) Claimed-recall scores—indicate the
percentage of respondents who recall
seeing the ad
(2) Related-recall scores—indicate the
percentage of respondents who accurately
describe specific advertising elements
32
Measures of Recognition & Recall
Many marketers reject recall as a valid
measure
1) Recall simply measures whether an ad
is received but not whether the message
is accepted
2) Recall is biased in favor of younger
consumers
33
Measures of Recognition & Recall
3) Recall scores generated by ads are not
predictive of sales performance
4) Day-after recall testing is biased against certain
types of advertising content
Understate the memorability of commercials that
employ emotional or feeling-oriented themes and are
biased in favor of rational or thought-oriented
commercials
34
Message Research Methods
Recognition & Recall
• Psychogalvanometer
• Pupilometer
Physiological Arousal
Persuasion
Sales Response
35
Measures of Recognition & Recall
Measures of Physiological Arousal
• Ads that are better liked are more
likely to be remembered and to
persuade
• Efforts are now made to measure
consumer’s affective and emotional
reactions to ads
36
Measures of Recognition & Recall
Measures of Physiological Arousal
• Galvanometer—measures minute levels of
perspiration in response to emotional arousal
• Pupillometric tests—measure pupil dilation
• Advertising researchers use changes in
physiological functions to indicate the actual,
unbiased amount of arousal resulting from
ads.
• Neuroscience – brain imaging tests
37
Message Research Methods
• Ipsos-ASI Next*TV
method
• Rsc’s ARS Persuasion
method
Recognition & Recall
Physiological Arousal
Persuasion
Sales Response
38
Measures of Persuasion
Used when an advertiser’s
objective is to influence
consumers’ attitudes toward and
preference for the advertised
brand.
39
Measures of Persuasion
Ipsos-ASI Next*TV Method
• Measures recall and persuasion of TV commercials
• Performs ad research in more than 50 countries
• Tests television commercials in consumers’ homes
• Tells consumers to evaluate a TV program, but
actually evaluating the commercials within the
program
• Mails 30 minute TV programs (like a sitcom) to a
national sample with embedded commercial breaks
40
Measures of Persuasion
Ipsos-ASI Next*TV Method
• One day after viewing—personal contact
with sampled consumers and measure
their reactions to the TV program and
the advertisements
• Measures message recall and
persuasion
41
Measures of Persuasion
Ipsos-ASI Next*TV Method
• Persuasion measured by:
Consumers’ attitudes toward
advertised brands
Shift in brand preferences
Brand-related purchase intent and
frequency
42
Measures of Persuasion
Ipsos-ASI Next*TV Method
1) Tests in a natural environment
2) Assesses the ability of TV commercials to
break through the clutter
3) Determines how well tested commercials
are remembered after this delay period
4) Allows the use of a representative national
sampling
43
Research Systems Company rsc
• Very active in message testing – tests TV commercials
in various stages of completion – rough cut to finished
form
• ARS schedules test sessions in eight metropolitan areas
• 800 – 1000 consumers from each metro randomly
invited to preview TV material with embedded ads
(including test ads)
• Before and after research design – basket of competing
products
• Tested over 40,000 commercials so far
44
Measures of Persuasion
The ARS Persuasion Method--rsc
1) Respondents indicate what brands they prefer
to receive among a “basket” of options—the
pre measure.
2) After exposure to a television program
respondents again indicate what brands they
would prefer to receive if their name were
selected in a drawing—the post measure.
ARS Persuasion Score = Post % for target brand –
Pre % for target brand
45
Measures of Persuasion
The ARS Persuasion Method--rsc
The ARS Persuasion score simply represents the
post-measure percentage of respondents
preferring the target brand minus the premeasure percentage who prefer that brand
A positive score indicates a shifted preference
toward the target brand
A higher-scoring commercial generates greater
sales volume and larger market share gains
46
Measures of Persuasion
The ARS Persuasion Method--rsc
ARS Persuasion scores are valid predictors
of in-market performance
The higher the score, the greater the
likelihood that a tested commercial will
produce positive sales gains when the
focal brand is advertised under realworld, in-market conditions
47
Predictive Validity of ARS
Persuasion Scores
Tested 332 commercials in seven countries ; involving 148 brands
representing 76 product categories; market shares before and after
advertising commenced were compared.
48
Message Research Methods
Recognition & Recall
• AC Nielsen’s
SCANTRACK
• IRI’s BehaviorScan
Physiological Arousal
Persuasion
Sales Response
49
Measures of Sales Response
Single-Source Systems
Gather purchase data from panels of
households using:
(1) electronic television meters
(2) optical laser scanning of universal
product codes (UPC) at retail checkout
(3) split-cable technology
50
Measures of Sales Response
Nielsen’s SCANTRACK
SCANTRACK collects purchase data by having
panel households use handheld scanners
Panelists record purchases of every bar-coded
product purchased regardless of the store
where purchased
51
Measures of Sales Response
Nielsen’s SCANTRACK
Use handheld scanners to enter:
Any coupons used
Record all store deals
Record in-store features that influenced
their purchasing decisions
Transfer data to Nielsen by calling a toll free
number and holding the scanner to the phone
52
Measures of Sales Response
IRI’s BehaviorScan
IRI knows what items each household purchases by
linking up optically scanned purchases with ID numbers
Members provide IRI with detailed demographic
information
Five markets in the US – communities that depend on
cable TV to get good reception – Pittsfield (MA), Eau
Claire (WI), Cedar Rapids (IA), Midland (TX), Grand
Junction (CO)
In each market 3000 households are recruited; approx
1000 are hooked to people meters
53
Measures of Sales Response
IRI’s BehaviorScan
Single source data consists of:
(1) household demographic info
(2) household purchase behavior
(3) household exposure to new television
commercials that are tested under real
world test conditions
54
Measures of Sales Response
IRI’s BehaviorScan
Splits cable signal before it reaches
panel households
Uses optically scanned purchase data
to know exactly which commercial
each household had the opportunity
to see and how much of a brand is
purchased
55
Measures of Sales Response
IRI’s BehaviorScan
Two types of tests are offered: (1) weight
tests and (2) copy tests
1) Weight tests—panel households are
divided into test and control groups and
frequency is varied
2) Copy tests—holds the amount of weight
constant but varies commercial content
56
Conclusions from TV ads
research
• What does it take for TV ads to enhance
sales and market share performance
– Ad copy must be distinctive
– Ad weight without persuasiveness is
insufficient
– Even good advertising wears out over time
– Advertising works quickly if at all it works
57
Conclusion 1: All Commercials Are
Not Created Equal: Ad Copy Must
Be Distinctive
• Commercials having strong selling propositions
are distinctive and tend to achieve higher ARS
Persuasion scores (e.g. Mentadent toothbrush
with a strong USP (flexible handle for gentle
brushing) scored 11.2 on ARS persuasion)
• Commercials for new brands tend to be most
persuasive, but commercials for established
brands can be made persuasive via brand
differentiation.
58
Illustration of a
Commercial with
a Strong Selling
Proposition
59
Conclusion 2—More is Not
Necessarily Better: Weight Is Not
Enough
• Ad weight means the number of gross
rating points (GRP) that support an
advertising campaign.
• An ineffective ad (not distinctive or
persuasive) has no likelihood of increasing
sales even if the ad weight is doubled or
tripled.
60
Relations among Advertising Weight, Persuasion Scores,
and Sales
61
The Relationship Between Media
Weight and Creative Content
• 47 commercials for established brands were tested and
classified as:
– Rational information (informative; argument based)
– Heuristic appeals (using celebrities, nice things like babies,
puppies, etc.)
– Affectively based cues (ads depicting warmth, love, affection,
companionship, friendship)
• Finding:
– Increased advertising weight led to significant sales increases in
sales only for commercials using affective cues
62
Conclusion 3—All Good Things
Must End: Advertising Eventually
Wears Out
• Advertising ultimately wears out and must
be refreshed to maintain or increase brand
sales.
• Familiar brands have been shown to wear
out more slowly than unfamiliar brands.
63
The Role of Sales-Effective Advertising for an Undisclosed
Campbell Soup Brand
64
Conclusion 4—Don’t Be Stubborn:
Advertising Works Quickly or Not at
All
• Some advertisers tend to “hang in there”
and wait for an ad to increase sales.
• Most of the sales impact occurs in the first
three months of a new ad.
• “Sunk costs” are an issue to consider, but
if an ad is not working at first, it probably
never will.
65