Advertising is often measured by its immediate effect – an Ad is published and it returned X dollars in sales. Sales divided by the cost of the advertisement is the immediate return.
However, there is a problem with only measuring the immediate return. In truth, past advertising has an effect on current sales. Additionally, past advertising has decaying effectiveness over time. In plain English, this means:
Past advertisements help current sales but only up to a point!
Measure Effectiveness of Past Ads
In 1979, the statistician Simon Broadbent introduced a model called Adstock. Adstock is a way to model advertising effectiveness while taking into account the delay of purchasing after seeing an Ad, and the decaying effectiveness of past Ads.
Advertising Carry-Over is the effect advertising has on a future purchase. Just because you saw an email Ad for a new pair of loafers doesn’t mean you’ll buy them today, but you may remember the brand next spring.
We can better understand the effects of advertising over time using the Adstock model. The graph below depicts a forecasted return on advertising with the number of advertisements displayed over time:
Within the first 10 quarters, Adstock reached over 250, while the Ad return average is less than $15,000. It’s not until 16 quarters of advertising that Ad return starts to spike above $15,000.
Adstock data can be used to understand the effect of previous advertising on current sales, and current advertising on predicted future sales.
Companies often measure advertising effectiveness using immediate return. As demonstrated here, accounting for past advertising is more accurate in measuring the effectiveness of advertising.
Using the Adstock model will help companies:
Save money by reducing advertising spending before diminishing returns.
Calculate how much money to spend on advertising for a specific return.
Identify which advertising channels perform the best by comparing Adstock returns.