The Importance of A/B Testing

Google does it. So does Amazon. Even the BBC does it. And there is no reason everyone shouldn’t be doing it.

It is A/B testing, and it’s an effective and unbiased means of understanding your customer’s habits and preferences.

What is A/B testing?

Understanding A/B testing is simple – you take something, a header image, perhaps, and you make two versions of it. You let part of your audience see the first version, and the other part see the second version. Then you look at how each version performed against your goal. At its core, it’s that simple.

For instance, say you have a weekly email that you send to customers and potential customers. For the moment you just want people to open the email and read the great content you’ve prepared, but in looking at your open rates, they are dismal. Why aren’t people even reading the email? Maybe they don’t recognize the email address it’s coming from, so they delete it immediately. Or perhaps, with the influx of email we get on a daily basis, it’s getting lost in the shuffle. Maybe your subject line just isn’t interesting. Using A/B testing, you can find out.

In the example above, we’d start by testing one of the suspected issues you’ve hypothesized as being the problem. Yes, just one – for now.  Let’s say it’s the subject line. Segment your list into two groups, and send the same email to both groups, but with two different subject lines. Which did better? Which got more opens? Was there any effect at all?

Testing email performance is easy and inexpensive, but testing elements of your website can be as well. Do you have a call to action that isn’t performing well?  Try testing different colors, placements of the CTA or wording- just remember to only test one variation at a time! Other elements to consider testing on your website include main navigation order, copy changes, images and layout. The most critical thing to keep in mind as you are planning your test is to consider, what are you trying to improve? That will help guide the variations you plan to test.

Based on your results, you can do more testing, tuning and fine tuning until you are seeing results that are more meaningful to your business, or can make a decision on marketing tactics based on those results.

It’s important to note that A/B testing isn’t new. In fact, it goes back to the heyday of direct marketing. But the testing time for a traditional, direct mail piece was much longer, taking months to segment your groups, mail the offer, evaluate the response and adjust. Digital follows this same process (segment, test, analyze, adjust) but can be done at a significantly accelerated rate – days or weeks instead of months.

So why is A/B testing important?

All of your investments in your business are important. In the old world of marketing, if you ran an ad in the paper for your store and it didn’t drive any traffic, would you have kept running the same ad? So why send out email, post-paid media or host a web experience that isn’t giving you the most for your marketing dollars?

A/B testing also removes any bias you or your marketing team might have. You may love the color green, for instance, but find that it’s driving your customers away. Or perhaps you feel that in your lead gen forms you have to get the name of the company the potential lead works for – but if you pull that field from your form, do you get an uplift of 2 or 3 times as many leads?

A/B testing gives you an unbiased, scientific means of understanding what will work best to keep your customer moving down the funnel, and now can be accomplished at the speed of digital marketing. So why IS A/B testing important?  Because it will help every dollar you invest in your marketing efforts go farther and work harder.