The key metric for a single TV ad (“spot”) is “impressions,” indicating the number of people who were shown the ad. This can be reported for all viewers or narrowed down to specific demographics defined by age, gender, socio-economic attributes, household composition, and more. Impressions for a specific demographic will always be lower than for the total audience.
Other common terms derived from impressions are ratings, TARP (Television Audience Rating Point), and GRP (Gross Rating Point). All three terms describe impressions as a share of the potential audience and are calculated by dividing ad impressions by the size of the audience owning a TV or using the subscription service.
Like impressions, these can be calculated for all viewers or specific demographics. Unlike impressions, the ratings for a specific demographic can be higher or lower than for the total audience.
In free-to-air TV, impressions and ratings are estimated based on a selected panel of households with people meters, which are electronic devices that record which channel is being watched and which household members are present. This panel data is then projected to the entire population, leaving room for statistical error.
In connected TV, the provider knows exactly how many times a spot was shown via return path data but does not automatically know who saw it.
Another term for impressions is “opportunity to see.” There is an important difference between an ad being shown (impression) and people actually seeing or hearing it. People often mute the sound, switch to other tasks, or leave the room during ad breaks. This happens less frequently for ads at the beginning or end of the break, which is why these positions are often sold at a premium.
Impressions and ratings are the trading currency of TV advertising. TV spot costs are determined by the number of impressions likely to be generated. Although future ratings are not known at the time of campaign bookings, media buyers can provide good estimates based on historical data. Depending on the contract, the station might guarantee a certain number of impressions for a given price, known as CPM guarantee contracts.
CPM, or cost-per-mille (thousand impressions), is calculated as the cost divided by so called equalised impressions. This metric is used to account for the fact that longer ads cost more, typically proportionate to their duration when the duration is an integer multiple of 30”. There usually is a small additional premium for commercials with odd durations, such as 50”.
A single TV spot is unlikely to have a meaningful impact, so a campaign consists of multiple spots. In addition to impressions and GRPs, campaigns are characterised by reach and frequency:
Reach is often discussed with a specific frequency range, typically open-ended on the maximum side. For example, an agency might be briefed to reach 50% of the target audience three or more times over a four-week period. Reach is measured using the same methods as impressions and faces similar issues, such as measuring opportunities to see rather than actual ad views and being subject to statistical errors.
Reach and frequency are important because they ultimately determine the effectiveness of a campaign, rather than just the number of impressions. Showing an ad to the same person ten times won’t necessarily increase sales, but showing it less than the required amount to make an impression won’t have any impact. The optimal frequency depends on the novelty of the product and the brand’s establishment, with lesser-known brands needing more repetitions.
Reach and frequency are metrics that are reliably measured by independent companies (e.g., Nielsen Media Research in many countries). However, they are not direct measures of a campaign’s success. A successful TV campaign works on multiple levels, and to fully appreciate its effects, all these levels need to be measured.