How a yes/no pop up with value proposition language impacts donor conversion during a high urgency campaign
Avila Foundation
Experiment Summary
Timeframe: 12/27/2021 - 01/04/2022
During calendar year end, Avila wanted to try an exit intent pop ups in lieu of a homepage takeover for the last few days of the campaign. There was concern that constituents who come to their site for content, may not want to contribute to a campaign. So we decided to test the language in the yes/no exit intent pop up. The control asked the site visitor if they wanted to donate to Avila to support their on-going efforts. While the treatment would have value proposition based language, asking the site visitor if they want to bring spiritual guidance to those looking for truth.
Site to their page was organic and driven by the blog content and resources they continuously provide to their constituents.
Research Question
We believe that adding a value prop based pop up for site visitors will achieve a higher donor conversion rate during calendar year end.
Design
Results
| Treatment Name | Conv. Rate | Relative Difference | Confidence | Average Gift | |
|---|---|---|---|---|---|
| C: | Control: Campaign driven language | 1.4% | |||
| T1: | Treatment: Value prop language | 3.1% | 124.6% | 94.8% |
This experiment has a required sample size of 558 in order to be valid. Since the experiment had a total sample size of 1,119, and the level of confidence is not above 95% the experiment results are not valid.
Key Learnings
The results indicate that Avila’s site visitors were more likely to give to their calendar year end campaign when presented with a yes/no pop up that identified Avila’s key value proposition. Without having any ask attached to the copy, that audience segment produced a 3.1% conversion rate resulting in a 125% increase in donor conversion rate with level of confidence of 94.8%. This tells us that Avila’s audience responds to their core mission and cause over just being asked for a donation. It is recommended that this concept is retested outside of calendar year end in order to determine if this is also true in a different context.