“I ran an incrementality test showing positive lift but didn’t see the results in the bottom line.”
Sound familiar?
I think we’ve all had this experience: we set up a nice, clean incrementality test to measure the effectiveness of a specific paid media channel. We get the results back: 5% lift, statistically significant. Nice! Champagne bottle pops, etc., etc.
Since we got this win, we bake the 5% lift into our forecast for next quarter and we sit back, ready to watch the money roll in as we scale spend in the winning channel.
But then, shockingly, we do not observe improved performance next quarter. And since we baked that 5% lift into our forecast, we’re in trouble.
What happened here?
The big issue here is that we didn’t consider uncertainty. When interpreting the results of our incrementality test, we said “It’s a 5% lift, statistically significant” which implies something like “It’s definitely a 5% lift”.
Unfortunately, this is not the right interpretation. The right interpretation is: “There was a statistically significant positive (i.e., >0) lift, with a mean estimate of 5%, but the experiment is consistent with a lift result ranging from 0.001% to 9.5%”.
The key point here is that when you’re doing any type of incrementality testing, you need to be looking at the uncertainty intervals from the test. You should never just report out the mean estimate from the test and say that it’s “statistically significant”. Instead, you should always report out the range of metrics that are compatible with the experiment.
When actually interpreting those results in a business context, you generally want to be conservative and assume the actual results will come in on the low end of the estimate from the test, or if it’s mission-critical then design a test with more statistical power to confirm the result.
If you just look at the mean results from your test, you are highly likely to be led astray! You should always be looking first at the range of the uncertainty interval and only checking the mean last.
