The Marketing Optimization Blog

Turning More Visitors Into Customers

Blog Home

Obama’s home page image test

By: Chris Goward
Date: November 10th, 2008

A couple days ago I wrote about how Barack Obama used Conversion Rate Optimization to maximize online donations.

Since then, one of our programmers, Matt Tobey, found more detail on what the home page script I found had been testing posted on schoonzie.com.

Here’s what we’ve found:

Variation A: Family casual black & white
The current page for first time visitors to the site looks like the image below. Since the test is no longer functioning, we can assume that this was the best-performing of the four variations that were tested.

Obama home page SwapBox location
Variation A: Red dash shows SwapBox placement

Variation B: Family waving black & white

Obama home page Variation B
Variation B

Variation C: Family waving full color

Obama home page Variation C
Variation C

Variation D: Obama and Biden couples full color

Obama home page Variation D
Variation D

In addition to the obvious image swap, the message below the images seems to vary slightly. It’s unclear whether this test was run as an A/B/n or a multivariate test so they could learn the interactions between the variables.

We still hope to interview the people responsible for the online marketing plan and learn more about the types of tests they were running. If you can introduce me, please email chris dot goward at widerfunnel.

Note: Don’t miss a post. Subscribe here for the WiderFunnel blog:


Posted in: Case Study

2 Responses to “Obama’s home page image test”

  1. Dylan Lewis Says:

    What do you think is the measure for success in this test? In the t-shirt version, it was probably t-shirt purchases - is it sign-ups?

    Is there a way to verify that it is sign ups?

  2. Chris Goward Says:

    The t-shirt offer test was actually on the donation page, Dylan, so I have assumed that the goal is a donation of any size, but I don’t have inside information to confirm that.

    For the home page test, there could be any number of goals that they could have tracked. If we assume that the primary purpose of the site was to raise campaign funds then the goal could have been a donation as well. But, they may have simply used a goal of reducing the bounce rate (ie. entering the site), or a combination of goals like donations, email signups, social media link clicks, etc.

    I’m hoping to get ahold of someone from the campaign that knows so he or she can share the learning with the eagerly interested community!

    Chris


Leave a Comment