Evaluating Return on Investment of Multimedia Advertising with a Single-Source Panel: A Retail Case Study
By Joan Fitzgerald of Arbitron, Inc.
"This study is important due to its methodology for it enables retail marketers to fine-tune their media strategies."
Source: Journal of Advertising Research, Sep/Oct 2004, 44(3), pgs. 262-270.
Type of Promotional Material/Activity Tested: Impact of radio and television advertising campaigns, and exposure to local and national newspapers and magazines, Sunday supplements, and retailer-specific free standing inserts on retail store visits and product purchase.
Sample Population: Arbitron’s National Marketing Panel (NMP) and Product Purchase Panel (PPP) each consisting of 500 persons, aged six and older, in roughly 250 households per panel. Participating households were selected at random from households installed in existing panels representative of Nielsen Media Research’s Philadelphia Designated Market Area.
Methodology: A pilot for a Portable People Meter (PPM) developed by Arbitron Inc. used to electronically track participant’s radio listening and television viewing as well as retail store visits. Readership data was collected via questionnaires. Product purchases obtained from Product Purchase Panel participants.
Independent variables: Radio, television, newspapers, magazines, Sunday supplements and free standing inserts.
Dependent variables:Shopping behaviors (e.g., day, time, and length of store visit and product purchases); Media consumption (radio listening, TV viewing, print media readership).
Top Line Results:
- A positive correlation exists between advertising exposure and volume of store visits. For example, shoppers with heavy advertising exposure in October indexed at 154, indicating their visits were 54% more frequently than the average shopper.
Index - Correlation of Advertising Exposure to Store Visits
|Heavy advertising exposure||154||124||123||134|
|Light advertising exposure||89||101||92||94|
|No advertising exposure||66||62||59||62|
Advertising exposure does not correspond with time spent in stores as it does with number of visits.
Weekday (10am-4pm) and weekend shoppers demonstrate distinct visit behavior. Weekday shoppers are less likely to shop on the weekend, while weekend shoppers are more likely to visit during the weekday too.
Distinct shopping behavior exists along demographic differences:
|Weekend Only||‘After Work’||Weekend||Weekday 10am-4pm|
Both heavy and light shoppers prove to be heavy media users (38% are heavy radio listeners, 42% are heavy broadcast television viewers, and 36% are heavy cable television viewers).
As an example of how retail shopping visits are driven by media advertising exposure (and their use as a viable tool to measure ROI in terms of store visit behavior), the study outlines the Weekend Shopper:
- Weekend shoppers are heavier radio listeners than the average shopper of a particular retailer
Broadcast TV ad spots shown Monday-Friday 4-7:30pm will reach 31% more weekend shoppers
- Radio ads played Saturday-Sunday 9am-1pm will reach 25% more weekend shoppers
- Cable TV ads shown Monday-Friday 4-7:30pm will reach 12% more weekend shoppers.
Take Away: This study is important due to its methodology for it enables retail marketers to fine-tune their media strategies. It provides “a research framework to evaluate the extent to which advertising exposure correlates with changes in retail store visit behavior. Using this framework in future research, marketers have the opportunity to modify their media mix and weight and measure the results to achieve greater returns on their media investment” -- including print advertising media in the marketing mix.
Complexity rating: 2 out of 3 (Complex statistical analysis scale: 1= none, 2= moderate, 3 = difficult)
Source: Evaluating Return on Investment of Multimedia Advertising with a Single-Source Panel: A Retail Case Study