McCann Erickson and Ad Promises Experiential Learning
McCann Erickson Worldwide Advertising is the world’s largest advertising agency
network. With offices in over 130 countries and more than 75 years of global experience,
McCann Erickson handles more global accounts than any other ad agency.
Joe Plummer is Executive VP and Director of Research & Insight Development for
McCann Erickson WorldGroup in New York. Recently, Joe has become concerned that
viewers may be overcome with the “dazzle” of television ads at the expense of actually
learning about the brands featured in such ads. With all of the special effects tools at the
disposal of advertising creators these days, what may become lost in the creative
execution of the television ads is the key selling idea for the brand.
To address the need for improved measures of viewer response to ads, researchers have
developed a set of multi-item scales called Ad Promises1. Eight-point Likert type scales
were used to measure each of the eight subscales of Ad Promises. The end-points of
these Likert scales were anchored by “strongly disagree” and “strongly agree”. The Ad
Promise scale items were prefaced on the viewer response questionnaire in the following
way: “For each statement, indicate your degree of agreement that the statement pertains
to the advertiser’s implied or stated promises about the brand in the preceding
A list of the Ad Promise scale items and other items used by McCann Erickson follow:
Ad Promise Scale Items
1. Problem removal
a. The brand would help remove a problem one may encounter
b. One would go from feeling annoyance to feeling relief as a result of using
c. Using the brand would NOT appeal to one’s desire for resuming the
pursuit of a goal.
2. Problem avoidance
a. Using the brand would enable one to avoid a potential problem.
b. The brand would dissipate fear and bring one a feeling of relaxation.
c. One’s desire for a threat-free pursuit of a goal would NOT be met by using
3. Incomplete satisfaction
a. The brand would bridge the gap between one’s expectations and the
existing circumstances in a situation.
b. The brand would bring an optimistic outlook to an otherwise disappointing
c. One’s desire for more complete satisfaction would NOT be met by using
4. Mixed approach-avoidance
a. The brand would take care of one of the remaining negative aspects of a
b. Using the brand would avoid impending conflict and bring peace of mind.
c. One’s desire for more consistency in thoughts about an object or issue
would NOT be met by using the brand.
5. Sensual gratification
a. A sample of the brand would make one want more.
b. Using the brand would take one from a neutral state to a pleasurable state.
c. The brand would NOT appeal to one’s bodily senses.
6. Intellectual stimulation
a. The brand would stimulate one’s intellect.
b. The brand would relieve boredom.
c. The brand would NOT appeal to my sense of adventure or risk.
7. Social approval
a. One would be considered more fashionable by using the brand.
b. The brand would help one feel less apprehensive in social situations.
c. The brand would NOT appeal to one’s desire for social approval.
8. Intrinsic satisfaction
a. Using the brand is its own reward.
b. The brand would be enjoyed for its own sake, not for what it will bring
c. Pure enjoyment of the brand would NOT be the only thing in it for
10. Marital Status
11. Children in house
13. Zip Code
14. Student Status
15. Formal education
16. Cultural Group
The Ad Promise scales were first evaluated by showing 413 consumers a 30-second
television ad for Mountain Dew in a lab setting and then having these consumers answer
the 24 items of the Ad Promise scales along with some demographic questions.
Some of the theory behind the Ad Promise scale items follows. Eight constructs are each
measured by three items. The third item defining each construct needs is reversed-scaled.
In other words, in analyzing the results, the reverse-scaled items need to be recoded. This
can be done by creating a new column in a spreadsheet where the total possible scale
points (Here, eight) receive an additional point (Here, summing to nine). Then, the value
for each respondent in the original variable to be recoded is subtracted from one-plus-the
total scale points). For example, in the current example, whenever a respondent recorded
a “7”, this would become a “2” after recoding (8 + 1 – 7 = 2). Likewise, in the current
example, whenever a respondent recorded a “1”, this would become an “8” after recoding
(8 + 1 – 1 = 8).
Go to the textbook’s website and download Ad Promises Not Reversed.xls from chapter
10. Create eight new variables corresponding to q1c, q2c, q3c, q4c, q5c, q6c, q7c, and
q8c. Name these q1cnew, q2cnew, q3cnew, q4cnew, q5cnew, q6cnew, q7cnew, and
q8cnew. Reverse code the original eight variables into these eight new variables. For
example, the first cell corresponding to the first respondent’s answer would be recoded
by subtracting the respondent’s answer from “9” (the number of scale points plus one).
The results for this cell could then be copied to the remaining 412 rows in this column for
q1cnew by clicking on the lower-right corner of the first cell and dragging down to the
row for the last respondent.
1. Recode the last item defining each Ad Promise construct (q1c, q2c, q3c, q4c, q5c,
q6c, q7c, and q8c) into new variables (q1cnew, q2cnew, q3cnew, q4cnew,
q5cnew, q6cnew, q7cnew, and q8cnew).
Save your Excel file as “Ad Promises Reversed Coded.xls”, and then import its contents
into SPSS in the following way:
Import into SPSS
Files of Type window (All Files)
Select “Ad Promises Reversed
Text Import Wizard Step 1 – accept default settings – Next.
Text Import Wizard Step 2 – variables arranged delimited – Yes;
– variable names at top of file – Yes.
Note: variable names will be taken from the first row of the spreadsheet.
Text Import Wizard Step 3 – accept default settings - Next
Text Import Wizard Step 4 – accept default settings - Next
Text Import Wizard Step 5 – accept default settings - Next
Text Import Wizard Step 6 – accept default settings – Finish.
Note: variable names will be at the top of the columns.
To make sure you do not inadvertently refer to the original variables, select “Data View”
tab of the SPSS window, and then select the top of each column of the original variables,
right click on the mouse and then select “Clear”. Now, the original variables are
removed from the matrix.
To understand how consistent each of the eight subscales was measured by its
corresponding three items, reliability analysis needs to be performed on each of the eight
sets of items.
2. Compute Cronbach’s Alpha for each of these eight sets of items.
Select the three items defining each construct in the
Reliability Analysis window
Put a check mark in the empty box “List
Composite variables can also be computed. In essence, a composite variable is a
combination of several variables into one new variable. Do this by using the Compute
command in SPSS in the following way:
3. Create a composite variable representing each of the eight Ad Promise constructs.
Creating a Composite Variable Retaining the Original Scale
Enter the name of the new composite variable in the
“Target Variable” box of the Compute window.
Enter the formula for the composite variable in the
Numeric Expression box of the Compute window [For
example, (q1a + q1b + q1cnew) / 3)].
Checking the Creation of Composite Variables
Select the composite variables you created and
move them into the “variables” box in the
Summarize Cases window by first highlighting
these composite variables and then clicking on the >
Click OK. You will receive a report on the
first 100 cases in the dataset. Inspect these
values to make sure they are all within the
original scale of 1 to 8.
1. Mark Peterson and Naresh K. Malhotra, “Measuring the Appraisal of Ad-
Based Affect with Ad Promises”, Journal of Business Research, vol. 42, (1998), 227-239.