Happy Election Year!
So, are TV metrics dead? No. We use them all the time to compare inventory across mediums.
Before the New Year, we used eCPM to compare the effectiveness of ad channels by removing impressions that weren’t actually viewed by voters in our target audiences.
My first job in political advertising was buying Facebook ads in the 2015–2016 cycle. Back then, digital video was only 12% of the $2.2B in political video spend. Digital always had two characteristics:
We know how many impressions were served
We price on CPM (cost per thousand impressions). Layer on targeting + viewability technology and you get eCPM.
The other 88% of video ad spend was on broadcast and cable TV where impressions aren’t as easily counted. Advertisers didn’t know exactly whose TV was on, let alone who was watching. Instead of using a CPM, they used rating points.
What is a rating?
Companies like Nielsen were started to help determine who was watching TV so networks could accurately price and sell ad space. No one wants to pay top dollar for ads on programs no one watches.
Nielsen built panels of people who filled out diaries (yes, pen and paper) of what they watched on TV. Nielsen extrapolated that out to calculate a rating for programs.
The higher the rating, the higher the price.
Panels are now supplemented with other viewership data, but the overall process is still the same.
How are program ratings calculated?
Define the total TV universe (18+, 35+, A25–54)
From past panel history, estimate how many TV households will watch
Divide viewing HHs by total TV HHs and multiply by 100
Example: 1.2M of 10M households watching = 12 rating
Higher ratings = higher prices for ads
What does this mean for buyers?
Panel methodology aside, on a program-to program basis, a rating is pretty straightforward. But a full TV schedule is made up of many programs, and simply adding up all of the separate ratings points tells half a story.
Total TV delivery is expressed as Gross Rating Points (GRPs) for the overall population and Targeted Ratings Points (TRPs) for a target audience. A typical political TV buy would deliver 1,000 TRPs per week.
TRPs = (Impressions to target ÷ Target universe) × 100
TRPs = 1,000
Audience = 100,000
Impressions delivered = 1,000,000
Average Frequency = 10
BUT, this does not mean everyone saw the ad 10 times. Exposure is duplicated if (when) viewers see your ad on multiple spots/programs. Some people saw 20, others saw 1. TRPs measure average weight, not reach.
What does this mean for political buyers?
Metric definitions matter - Know exactly what is being reported. Is performance against target voters or the general population? Don’t mix universes.
Measure TV - TRPs reflect weight, not exposure. They do not show how many times a voter actually saw an ad. Use independent measurement to manage frequency and avoid waste.
Compare mediums apples-to-apples - Use common currencies (TRPs, eCPM, incremental reach) and compare channels on the same basis. Shift dollars to what delivers the most verified voter exposure.