Monday, April 25, 2011

Lessons from media in understanding consumer behavior

What do gross impressions and average frequency have to do with shopping multiple stores or using multiple sources in decision making? Not much, but you can use the formulas developed for calculating average frequency to calculate average number of different stores shopped, average number of media sources used for decision making, etc. If you have access to consumer survey results, you can calculate these and identify behavioral trends.

Calculating average frequency in the media industry has been around for quite some time. See below for basic formulas and Exhibit 1 for media schedule example:

Gross Impressions/ Total Unduplicated Reach # = Average Frequency


Gross Rating Points (%)/ Total Unduplicated Reach % = Average Frequency

Media Calculations (from Exhibit 1):

Gross Impressions: 92,000 + 87,000 + 65,000 + 99,000 = 343,000

Total Unduplicated Reach #: 123,000

Average Frequency: 343,000/123,000 = 2.7886


GRPs (%): 7.7% + 7.3% + 5.4% +8.3% = 28.6%

Total Unduplicated Reach %: 10.3%

Average Frequency: 28.6%/10.3% = 2.7886

This frequency calculation has value beyond the media world. In Exhibit 2, I’ve demonstrated how you can calculate the average number of different stores shopped in the past 7 days.

Shopping Calculations (from Exhibit 2):

Gross “Shopping” Impressions: 744,000 + 690,000 + 450,000 + … = 3,636,000

Total Unduplicated “Shopping” #: 1,160,400

Average number of different stores shopped: 3,636,000/1,160,400 = 3.1334


Gross “Shopping” Rating Points (%): 62% + 58% + 38% + … = 303.0%

Total Unduplicated “Shopping” %: 96.7%

Average number of different stores shopped: 303.0%/96.7% = 3.1334

What can these numbers tell us?

I’ve found them effective in demonstrating competitive shifts and consumer trends. Notice in Exhibits 3 and 4 that the overall percentage of households in Any Metro USA having shopped any grocery stores in the past 7 days hasn’t changed, yet the average number of different stores shopped has increased 23% in the last five years. You can go a step deeper and apply this method by store to see which store’s shoppers are most likely to shop multiple stores and how those trends compare to the competition. In essence, this can provide a measure of store loyalty versus the competition and over time.

If you have research questions that allowed the sample to select more than one choice among multiple, then you have the ability to make these calculations. You can calculate average number of different stores visited or shopped, web sites visited, sections read, sources used in decision, items bought, etc. This will allow you to conduct analysis where you may have thought you hit a dead-end.