The relationship between market revenues and the corresponding variance in cost-per-point ranges by month. There is a strong correlation (within one standard deviation, or said another way, within the same sampling error as in your ratings service company) between monthly total market revenues and cost-per-point levels for these same seasonal periods.
It has been found that for young adult demos, cost-per-point ranges could be forecast to within +/- 5% by annualizing the monthly market revenue figures, multiplying that number by 70% then omitting the last six digits of the number, i.e. dividing the result by 1,000,000. The formula is (12Y x .70) - 1,000,000 = CPP Forecast +/- 5%, where Y is the total monthly market revenue figure, and .70 is the Market Revenue Index Factor.
For example, if the monthly total market revenues for a specific market was $10,000,000 then one can be 95% confident that the CPP (average cost-per-point) range for that month is $84 +/- 5%, or between $80 to $88. The average cost-per-point range takes into consideration clearance and preemptibilty issues. When stations take lower CPP business resulting in preemptions and oversell, the actual CPP of the schedule rises. Having a method to forecast average cost-per-point ranges by season (as well as better yield management) would save millions of dollars to both advertisers and stations alike, less preemptions, oversell, and better performance.
Indeed, a cost-per-point calculation should reflect future conditions not historical ones. That is the fallacy with advertising agency cost-per-points. There is no accounting for how the future differs from the past. Additionally, the study examined station's with established branch franchises, and how these differentiated stations compare with the above formula. Stations, ,or programs that are highly differentiated among their competitors would utilize a higher market revenue index factor.
Also, market growth and dynamics play an important role. If the market has consistent double digit growth, then the market revenue index factor should be adjusted as well.
