A Dumb Way To Buy Media Is Based On The Cost Per Thousand People Exposed—CPMs.

Why do companies by media in terms of CPMs? Because that’s the way it’s been bought for fifty years and who would argue with tradition?

Television advertisers have consciously or unconsciously assumed that, for the most part, a viewer is a viewer is a viewer.  The only number that counted was the total number of 18-to-34 women, or 24-to-44 men (or some other simple demographic) watching a given program. 

But if, as research has shown, program involvement enhances advertising response, and if involvement means more than simple viewership, then the cost per thousand people involved (CPMI) is a better tool than cost per thousand exposed (CPM).
Media vehicles do differ considerably in terms of involvement (RE: engagement).  An advertiser that understands this difference can use it to improve a campaign’s success.  The question is: How closely related are the CPMI and the CPM for the same programs?  Would a company make the same media buy if the decision were based on impacted viewers rather than the total number?

The answer turns out to be “no!”  We have discovered three important things  First, two programs with the same ratings and the same type of audience demographics don’t necessarily have the same levels of involvement.  Since one program maybe more involving than another (audience size and composition held constant), one program will be more effective than another in terms of advertising response despite comparable costs.

The second finding follows from the first.  If you start with a fixed advertising budget (say $50 million) and buy a media schedule based on cost per thousand (CPMs), then do it again based on cost per thousand involved (CPMIs), you end up with two different media schedules. 

Third, and most importantly, the estimated advertising effectiveness of the CPMI buy is far greater than the CPM buy.
Everything we’ve done suggests that CPMIs will be the wave of the future.