Because SmartServe is an ad server, it can randomly assign a fraction of listeners to not receive the ad, ie to be in the control group. This control group should be close to identical in all ways to the exposed group, with the only difference being that the exposed listeners received an ad in this campaign. We can then compute attribution across for both groups, with the difference in purchases between both groups being the causal impact of the campaign.
The two main options are “ghost” ads and PSA ads:
Holdout Type | How It Works | Accuracy | Opportunity Cost | Financial Cost |
“Ghost” | When Podscribe’s ad server is called, a listener is randomly assigned to the holdout or exposed group. If holdout, SmartServe’s VAST server returns a blank response to the publisher’s ad server, and if exposed, it returns the real ad.
The main limitation is that the publisher’s ad server may still impose some light filtering on the “exposed” group, that the holdout group is not exposed to. This is typically minor, and can usually be quantified with the Podscribe reliability score. Podscribe is working on a workaround to solve for this to match PSA accuracy. | B+ | Low - Moderate | None |
PSA | Instead of not serving the holdout group any ad, Podscribe serves holdout listeners an “ad” that is unrelated to the brand.
Because the holdout ads are actually served, this 1) necessitates payment for these impressions and 2) increases accuracy because both holdout & exposed listeners are subjected to the same exact selection process, end-to-end. | A | Low - Moderate | High |
Coming Soon: How to set these up and view results in the dash
Truncated IP Measurement (beta)
In a limited capacity, Podscribe can support truncated IP attribution. Through SmartServe, Podscribe can create control and exposed groups of truncated IP listeners, using either the ghost or PSA methods above. Then, for each group we can compute how many total conversions occur on all masked IPs behind the truncated IP. Eg if the truncated IP is 10.1.2.0, then all conversions on IPs starting with 10.1.2. would be counted. Aggregated attribution counts will then be available like usual in the Podscribe dashboard, although these conversions will not appear in the Individual Conversions report, since we are able to deterministically link conversions with this approach.
This may not work for a variety of reasons, most specifically noise, or if the campaign cannot be SmartServed. We advise keeping expectations low, and always still requesting non-truncated IPs, as that will always be more accurate. For this reason, we recommend keeping expectations low.
Since this solution requires attribution to be done on a significantly higher amount of IPs, there likely is an additional fee for truncated IP attribution. Please speak with partnerships@podscribe.com.
Powered by