The Numbers Game

TV and radio have two standard ways of presenting audience — as a whole number and as a percent of a market’s population. The first is gross impressions, which is used in CPM; the second is gross rating points, which produces cost-per-point.
The recent industry proposal, that spot media make things easier by switching from a cost-per-point to a cost-per-thousand currency measure, ignores important differences-and has competitive implications well beyond the numbers game.

Cost-per-thousand is based upon impressions. Cost-per-rating-point is based on coverage. The big difference is a rating point counts only impressions delivered to a specific area, where a CPM counts all impressions.

Converting a CPP to a CPM is simple. You just have to know how many thousand viewers or listeners are in a rating point. For example, if a market has 1 million adults 18-plus, then a rating point is in-market impressions equal to 1 percent of that, or 10,000. If the cost of buying that rating point is $100, then the CPM is $100 (10,000/1,000), or $10. It’s an easy calculation if you know the population of the market on which the cost-per-point is based. This is readily available.

Changing a CPM into a cost-per-point is more labor-intensive. You have to know how many of the medium’s total impressions were delivered to the desired market area. Cost-per-point is the right currency for spot.

National media traditionally use CPM. Spot media’s preference for cost-per-point isn’t an accident. Spot is market-specific and cost-per-point emphasizes its ability to reach specific high-value areas. Cost-per-point also simplifies spot planning. If the brand needs 100 adults 18-plus points a week in a market to achieve a 40 weekly reach goal, the buyer takes the CPP, which is $100, and multiplies it by 100 to get the weekly schedule cost of $10,000. You can’t do that with cost-per-thousand unless you convert it to rating points and that requires defining a market area.

Which is the better measure?

The better measure is decided by the media plan. Cost-per-point is the cost of geographically targeted impressions and is most useful for buying media like radio and spot TV.

Cost-per-thousand is the cost of impressions delivered anywhere and is most useful for buying national media like network and cable television.

But the two measurements have competitive implications. Media with dispersed audiences like cable TV will look much better on CPM than on CPP. Their geographic area is large so their audience as a percent of the population is small. That means a rating point costs a lot.

A spot medium like radio will do better with CPP than CPM because its impressions are concentrated in a specific market area. And the many radio impressions delivered outside that area, which would be counted in CPM, are often not measured.

Fifty years of TV as the most important medium in dollars has had an effect on media systems. In the name of efficiency, we would like to data-input everything as we do TV. Even our planning models, which focus on integration, harmonization and comparability, seem to suggest one measure fits all. CPM doesn’t.

The TV network model is national; other media, like outdoor, newspapers and radio, are local. Advertisers are willing to pay for the geographic selectivity that targets a brand’s market. Cost-per-point is the best measure of that.

Erwin Ephron is a partner of Ephron, Papazian & Ephron, a leading consultancy to advertisers and the media industry. He can be contacted at