We already know that, just like in science, you should always have a control before having a variation. Before you introduce clients to a new way, make sure there’s a tried and tested “old” way. On that note, don’t disrupt your existing users with seemingly unstable modifications. They won’t like that. Rather, run your A/B tests on new users and make the overall changes once you have concrete data in preferring one method over another.
Consistency is also often overlooked in the A/B world. If you’re changing one dimension, then that element should carry across your entire campaign. A graphic or color change, for example, should be reflected throughout your site and landing pages. Still, there’s a whole world of A/B tips that doesn’t even get explored. Consider product images for starters. Try going beyond just plain product images shot in shoddy light with your iPhone and no editing. As great as your iPhone is, it still can’t really capture details and light the way a professional can. Keep a professional on call and watch your conversion sky-rocket. You can also use the images to limit the dreaded dead end of conversion, it “shopping cart abandonment.” Move beyond it by “following” the customer. Send an email reminding them of their shopping cart (including the beautiful product images). Little things like how you phrase “shopping cart” make the difference. It’s not just a shopping cart, it’s your basket items they wanted.
Segmentation. People often make the mistake of thinking A/B testing is an all or nothing measure. Entrepreneur and self-proclaimed conversion optimization junkie, Peep Laja understands that “An experiment that seemed to be performing poorly might have actually been successful, but only for a certain segment. For example, our experiment may have shown that a variation of our mobile landing page is not performing well. When looking into the segments though, we may see that it is performing exceptionally well for Android users, but badly for iPhone users. When not looking at segments, you can miss this detail.”
Laja recommends three segmentation strategies that include dividing users by source, behavior, and outcome. Source, for example, looks at where people came from, ie. email campaigns, Google, Facebook, Twitter etc. Once you’ve determined this, study the various conversion and bounce rates among the segmented user population. You can the data to either strengthen your weak spots or focus on giving more attention to successful platforms.
As far as behavior goes, Laja encourages businesses to look at user needs or goals. Are people coming to your site for one specific need, searching for one or two specific items? Is this segment group within one region or spread across multiple regions? Lastly, Laja says to “Separate people out by the products they purchase, by order size, by people who have signed up, etc. Focus on groups of people who have delivered similar outcomes and ask questions like the ones above.”
At the end of the day, A/B is more about human relations than it is about science … at least it should be. You should also always be engaging your customers. Talk to them. See what they want. Understand their experience. This is the real A/B and it’s where you benefit from the gift of others’ experiences. It’s about qualitative data over quantitative data. You can always offer your customers surveys and reward them with freebies and discounts. These offers should always be fair trades. The more involved the survey, the better the reward. At the same time, set real metrics. Understand what your benchmarks are, what’s realistic, and where you want to be. A/B testing is never just about seeing what people want. It’s about refining your business practices and creating a better user experience.
Are you ready for a smarter way to engage with your customers?
Benchmark helps you do email marketing the practical way. Create an ongoing relationship with your subscribers that leads to increased sales and happier customers.