Paid search data

Google Query Volume (GQV) is often an important confounder between media and sales. This is particularly true for paid search because query volume can drive ad volume under certain campaign settings, such as when the budget cap does not prevent it. When GQV is a confounder, you must control for it to get unbiased causal estimates for any media it confounds with. Failing to control for GQV can lead to overestimation of the causal effect of paid search.

Meridian offers a simple solution in that GQV data can be included as a control variable. Consider the following recommendations:

• It is best to scale GQV data by geo population. This can be done with the control_population_scaling_id argument.

• Paid search campaigns that target brand-specific queries are very different from those that target more generic product related queries. It is best to include these campaigns as separate media channels in the model.

• Brand-specific keyword campaigns are often modeled using clicks because they are intended to drive direct web traffic. Other campaigns are often modeled using impressions because impressions can be effective even when they don't result in a click.

• It is best to include query counts that correspond to the queries targeted by each paid search media channel included in the model. For example, if the MMM includes Brand-specific paid search and Generic paid search as separate media channels, then it is advisable to include Brand-specific GQV and Generic GQV as two separate control variables.

For more detailed considerations, see Including query volume as a control variable.

GQV, paid search, frequency, and other Google-related data for your organization can be obtained from the Google MMM Data Platform. For information about accessing this data, see Use MMM Data Platform.

Using clicks or impressions for search ads

When you are deciding whether to use clicks or impressions for search ads, consider the following:

• Deciding between clicks and impressions does have an effect, but there is no consistent perspective across, or even within, measurement partners on which input is best.

• As an advertiser, consider what you have more control over. Because the model tells you how media execution (whether defined by clicks or impressions) impacts a KPI, using a variable that you have more control over, in turn, gives you more control over impacting the KPI.

• Clicks are more likely to be correlated with the KPI, while impressions are more likely to be correlated with GQV. Brand-specific keyword campaigns are often modeled using clicks because they are intended to drive direct web traffic. Other campaigns are often modeled using impressions because impressions might be effective even when they don't result in a click.