Digital analytics have come a long way in the past couple years. More reliable tracking mechanisms, advanced software, and broader understanding of how to measure the impact of channels has made us all smarter. However, we’ve also become guilty of occasionally jumping to conclusions too quickly now that we can see performance by channel on a minute to minute basis. We may inject spend into test cells that appear to be performing well only to find what appeared to be a great pocket of performance was actually a statistical blip. Or, we may pull money out of seemingly under-performing channels and discover overall program results plummet because cause and effect correlations were not taken into consideration. There are many victims of overzealous budget re-allocation but Display Prospecting tends to be one of the most popular.
Don’t get me wrong, I love seeing the high ROIs from dynamic remarketing and the low CPAs from advanced search strategies boost overall results. But, I try not to forget where the qualified volume came from to support a successful remarketing program or where the awareness came from to deliver branded search results.
We recently ran a series of controlled tests within the retail space to quantify the impact channels have on one another. One of the most interesting results emerged when we looked at how display prospecting affects remarketing and paid search.
We funneled 3x more spend into Display Prospecting for three weeks in a targeted DMA while keeping remarketing and search budgets constant. We then compared KPIs in the DMA against total campaign benchmarks pre and post-test.
After three weeks, an injection of spend in Display Prospecting led to:
- 70-90% increase in remarketing and search conversion rates (see figure below for pre-test and post-test index comparisons against a campaign baseline of 100)
- 25-30% decrease in CPAs
The Bottom Line
These kinds of experiments are easy to conduct (and should be conducted) in your own marketing campaigns. The results can then be incorporated into cross-channel planning models to fully harvest the effects each channel has on one another.