Account-based marketing (ABM) shows a tremendous amount of halo effect
Paid ABM initiatives may not show a lot of engagement at first when looking at traditional click-based attribution, however, a causal impact analysis for a leading SaaS advertiser showed that performance was significantly greater when targeting certain companies through LinkedIn vs. companies that were not targeted. The delta between groups was vast, and the results were statistically significant.
Experiment Results
A first experiment took place in Jul-Oct 2020: We compared lead volume of targeted companies versus a control group of non-targeted.
Lead volume across test companies directly associated with ABM-specific efforts based on UTMs was 120. This is bad performance for what we spent.
Lead volume across test companies beyond UTM-based attribution was 760, meaning that 677 leads were produced across targeted companies without explicit UTM tracking.
Therefore, the halo effect was 9.2x from 83 up to 760.
The probability of obtaining this effect by chance is very small (p=0.003 based on the Causal Impact method), deeming this causal effect statistically significant.
When counting the non-click attributed increase in leads this campaign became very profitable.
A second experiment took place in Nov-Dec 2020 using a different ABM list, and showed similar trends although not as drastic as the first experiment:
Lead volume across test companies directly associated with ABM-specific efforts based on UTMs was 83.
Lead volume across test companies beyond UTM-based attribution was 367, meaning that 247 leads were produced across targeted companies without explicit UTM tracking.
Therefore, the halo effect was 3.1x from 83 up to 367.
The probability of obtaining this effect was very small (p=0.002 based on the Causal Impact method), deeming this causal effect statistically significant.
ABM is often undervalued
These test results mean that:
If you’re running any ABM effort, it is very likely that a large portion of unattributed leads in your CRM are in fact the result of your ABM initiatives!
If you’re not running any ABM effort currently, then you might want to run a test in order to quantify any delta between actively targeted accounts vs. everything else using a more sophisticated method - beyond traditional UTM-based attribution.
About the Causal Impact method and the plots
The Causal Impact R package implements an approach to estimating the causal effect of a designed intervention on a time series.
The above plots contain three panels:
The first panel shows the data and a counterfactual prediction for the post-treatment period.
The second panel shows the difference between observed data and counterfactual predictions. This is the pointwise causal effect, as estimated by the model.
The third panel adds up the pointwise contributions from the second panel, resulting in a plot of the cumulative effect of the intervention.
See: https://google.github.io/CausalImpact/CausalImpact.html
.