💭 “Would I get more conversions if I use a blue CTA that’s more on-brand, or a red version that stands out more on the page?”

It’s a question that, at some point, anyone with an online business has asked themselves. Split-testing is nothing new to the e-commerce world: in other areas of the market, it also goes under the name of A/B testing and has been a tried and tested tool for businesses to see what converts best, and what doesn’t.

In short, split testing helps you make decisions based on data, rather than guesswork.

iPaper’s URL Split-Testing tool essentially brings this functionality to your Flipbooks, allowing you to test two separate Flipbook URLs against each other, and measure how your customers react to each.

If you’re not already familiar with what Split-Testing is, and how it works, read our guide on why you should be split testing, and how to set it up.

When split-testing, it’s important to do so sequentially, in order to understand exactly which change has yielded a desired result. If you test too many variations at the same time, you might still achieve a positive result, but it’ll be incredibly difficult to pinpoint, and replicate exactly what facilitate the result.

So if you’re wondering whether a certain color CTA converts better than another, and whether displaying 10 items on a page instead of only 7 converts better, make sure to test one thing at a time.

If your proposed test means exposing your customers to anything more than a control (version A) and a variant (version B) of your Flipbook, this will work best when you test sequentially.

Essentially, you’ll only ever test one variant, versus the control, at any time.

The ‘winning’ variant of each test in your sequence will form the basis for your control in the next test in your sequence. That way, you can trace exactly which changes have yielded positive results, and which haven’t.

While you won’t be able to define your sample size through iPaper’s Split-Testing tool directly, in order to collect reliable data, you’ll want to expose as many of your customers as possible to your tests.

As the only customers who will be exposed to your split-test are those who access it via the custom URL generated by the Split-Testing tool, make sure that this link gets shared to as broad a test group as possible. This way, you’ll ensure you get results that accurately represent your audience.

When testing different variants, it’s just as important to have a defined test period as it is to have defined test criteria. For example, if you’re using a holiday season to test whether a certain presentation of your products works better than another, arrange for your test to begin in the run-up to the season, allowing as many customers to see your Flipbook, and stop your test promptly after the season finishes.

This way, your results will provide you with a snapshot of how your variant performed during this finite holiday season.

Like ensuring that your tests have a defined start and finish, it’s important to not make changes once you commence testing. By creating changes midway, you’ll disqualify all the data gathered prior to making your change! If you need to make a change, it’s better to stop your test, make your required changes, and start a new test, to ensure the integrity of your data.

In many cases, you'll want to understand what effect your split testing has on your conversion rate, i.e., whether one CTA leads to more hits to your e-commerce store than others.

In this case, consider using UTM parameters to track how many visitors have been referred by a specific Flipbook.

Get started with our guide on how, and when, to use UTM tags in your Flipbooks, below:

With our best practices on how to leverage our URL Split-Testing tool, you’re ready to test away and let data inform your Flipbook changes.

Of course, if you're unsure of anything, or have any questions about split testing your Flipbooks, reach out to us using the chat bubble in the corner. ✌️