Measuring Incrementality with Facebook

Performance marketers are often measured against the effectiveness of marketing dollars. A fantastic way understand advertising effectiveness is through measuring incrementality.

Incrementality is the measurement of impact that can be reasonably attributed to advertising spend.

Measuring these effects of advertising remains quite challenging even with granular level data. In recent years, Facebook has brought incremental lift tools across conversion and brand. You can combine the two for what Facebook is calling a Power Lift Study. They can be applied account wide or to specific campaigns.

Because I’m a performance marketer, I’ll focus on conversion lift. The design involves taking two groups and showing ads to one and not showing ads to the other. The group that will see ads is the test. The group that will not see ads is the control. Over the course of time and investment, you’ll be able to measure how much your business benefits from ad exposure in the test vs control.

It may help with an example. Let’s say we have an audience of 1M. We can run the experiment and tell Facebook we’d like 10% to be the control (100k) leaving the remaining as the test.

After time and investment we can measure the difference in activity between both groups. If the test group spent 50% more than the control we can reasonably say the advertising “lift” can be described as 50%. Statistical power is important to measure the confidence in these results. Per Facebook anything above 75% is considered statistically significant and if the tests were to be re-ran you would likely achieve the same results.

These tests are readily available and self serve in the “experiments” section of Ads Manager.

Happy testing!