Recently I read an article in WIRED (May 2012 issue): The A/B Test: Inside the Technology That’s Changing the Rules of Business by Brian Christian, that led to some new thoughts about making the most of your website.
The example illustrating the article featured the Clinton Bush Haiti Fund, showing the increases in online fundraising as a result of A/B testing of various segments on their webpage.
A/B testing, in case you don’t remember from your marketing class, is a simple way to test changes to your web page (or direct mail letter, etc.) against the current design and determine which ones produce the best results. By testing your changes before implementing them you get quantitative answers about what modifications work best. Testing and tweaking can increase donations, responses, or whatever else your goal may be.
You may have tried A/B testing with direct mail pieces in the past. You may have wondered if the costs, effort, and time of test mailings were really worth the effort. But, today, you need to look at the issue again, with respect to your website, and realizing that the game has changed greatly with the latest advances in technology.
Some of the results the Clinton Bush Haiti Fund gained as a result of A/B testing were:
1. They began with a website page with one column of text and no picture. When they added a photo, dollars donated per view increased 1%. When the page was redesigned into two columns, including the photo, and placing the response form side by side with the text, they saw an 8% increase!
2. Increasing the size of the font on the response form led to a 2% increase
3. Reducing the number of fields on the response form (eliminating title, phone number) netted an 11% increase.
4. Deleting the “security” icon reduced dollars per view by 5%.
4. Changing the language on the “donate” button from “submit” to “support Haiti” resulted in a 16% increase in dollars donated per view.
Now, nobody can promise you similar results. But, this example makes an important point. No matter how brilliant and intuitive your creative staff are in determining what will be most effective on your website, using real data is better. (I enjoyed Christian’s explication of how we usually work: HIPPO—relying on the Highest Paid Person’s Opinion.)
Nor do we all have budgets that will support the purchase of apps like Optimizely which can take the drudgery out of constructing these tests.
But why wouldn’t we make use of these ideas and test and tweak some of the critical items on our websites in order to improve results! You don’t need to take big risks to test one headline this week and another next week. Why not use real-time data to make website decisions? Not much to lose.
One caution Christian makes: Don’t make hard and fast rules from the conclusions of your A/B testing; it isn’t meant to discern great principles. But it does give you quick, data-based tools you can put to work immediately! Think about it.