Politics aside, as a marketer it’s hard not to be in awe (and a little jealous) of the work the Obama Digital team pulled off on the Obama Change campaign with A/B testing. Imagine the excitement in the air as you and your team has been working hard for months on a campaign you feel passionate about, and then to see all of your efforts help increase the donation conversions in real time, and ultimately help raise $690 million in fundraising by increasing donation conversions by 49% and sign up conversions by 161%.
But before we look into greater detail what exactly how the Obama Digital Team used A/B testing to achieve this feat, we should probably first look at what A/B testing is:
In essence, A/B testing is an experiment where you compare two slightly different designs of an element, which can be anything from a website, a magazine ad, or even a product label, and see which version is more successful. Version A is usually the original design, and version B is the one with the slight change. If we were testing two different designs for a website to see which format performed better, we would redirect half of the website traffic to go to Version B and then measure (using nifty tools like Google Analytics, Optimizely, and Blue State Digital tools) metrics like number of sales or the bounce rate for both versions. Using these results we can then optimize our website and use the best performing version to increase online sales.
So back to the Obama Digital team – over a 20-month period and over 500 a/b tests later, almost every element of the Obama website had been optimized. The team used a/b test results to create a simpler donation format:
Which the Romney campaign later copied:
They also found which images and where those images were placed would appeal most to viewers:
By taking advantage of A/B testing, marketers are able to see how simple changes in color, page speed, website design, images, and copy have an effect on the viewer’s behavior.
Now most of us don’t have the sexy life of being a digital marketer for the President’s election campaign, but that doesn’t mean we can’t conduct our own A/B tests, use the same measurement tools, and see some exciting results for our own company website.
Want to learn more about how A/B testing was used successfully for the Obama Campaign? Read Kyle Rush’s personal experience working on the team: http://kylerush.net/blog/optimization-at-the-obama-campaign-ab-testing/
Other good reads: