Description: This article outlines how InnovidXP deals with seasonality in our attribution method.
The impact of TV advertising is isolated by using control groups to estimate a baseline. Without a baseline, if you advertise in the run-up to the holiday season, it would be impossible to ascertain whether an increase in traffic is due to advertising performance or simply the fact that interest is naturally high during this period.
When calculating attribution, we estimate how many visits would be expected in the absence of advertising (the baseline measure). The estimate comes from the visit rate of a group of households who have not been exposed to advertising. This control visit rate is calculated daily. During periods of high interest, the control visit rate will be higher, the number of expected visits will be higher and we only attribute visits from the exposed group that are over and above this seasonal baseline.
The factor used to upscale attribution from the panel should not be impacted by seasonality. When web traffic rises due to seasonal interest, we would also expect to see sessions captured in the panel increase, and the ratio between them remain stable.
What is a control group?
A control group within the scientific field is a group separated from the rest of the experiment so that an independent variable being tested will not influence the results. This approach is used in the InnovidXP platform, where control groups are used to estimate the expected number of visits or expected (i.e., unexposed) visit rates to a website or app. For example, control groups are used to calculate the Response Percentage metric in InnovidXP outcome reports.
How InnovidXP control groups work
|The image on the right shows a spread of households. The exposed group, represented here by the blue dots, is composed of households exposed to a brand’s TV spot in the week it aired, or in the 30 days prior. The remaining green dots have not been exposed and so are eligible to potentially be added to a control group (or unexposed group).
How control groups are used
The number of visits captured from the control group can tell us how many visits we should expect following the airing of a TV advertising spot (of a given size).
The difference between the actual number of visits and the expected number of visits forms the incremental attributed response. Daily control visit rates can vary due to seasonal and weekday patterns. For example, Cyber Monday would show different volumes compared to an average day.
Incremental attribution with control groups
Why do advertisers use control groups?
Broadly, using a control group allows advertisers to assess the effectiveness of an ad campaign by comparing the responses to those exposed to advertising with responses to those who are not exposed. Using a control group allows advertisers to account for external factors that could impact results, such as seasonality. When control groups are enabled, responses are attributed only when they exceed control.