The Top Five Toughest Items to A/B Test on E-commerce Sites (That Should Be Tested, Anyway)

No online marketer worth his or her salt should undervalue A/B testing for e-commerce sites. After all, these sites always aim to convert visitors into buyers, and one of the most efficient ways to improve their conversion rates is through A/B testing different elements.

Also known as split testing, A/B testing is a way of testing multiple versions of an element to see which one converts better. For example, half of your visitors will see landing page A, while the other half will be directed to landing page B. These two versions are then compared and analyzed to see which one performs better.

Ideally, you should test every single element on the site to optimize conversion. Unfortunately, the following items are a little bit harder to A/B test, although doing so can reap substantial rewards:

1. Pricing

Product prices are one of the key elements of every eCommerce website, so one would think testing them should be fairly simple. But while there’s usually no problem with presenting different prices, it’s tough on most eCommerce platforms to ensure that the system charges only what is shown on the product page according to the test variation. Plus, it’s extra tough for international sites that transact in multiple currencies.

2. Shipping Strategies

When you have to deal with different shipping options like FedEx, UPS or USPS, it can be difficult to A/B test all of them. Some eCommerce sites also make it almost impossible to bring the shipping estimates to the product detail page or shopping cart, where it is convenient for most customers.

Free shipping promotions that are done in exchange for, say, subscribing to a newsletter or signing up for a club, can also be hard to test on most conventional A/B testing tools due to the complexity of the process. The same goes for shipping discounts once the consumer reaches a certain amount in his shopping cart.

3. Checkout Process

Many eCommerce platforms have hard-coded checkout processes that make it impossible to change the steps involved. So, if you want to reduce the number of steps during a consumer’s checkout process (or create a single-page checkout), you might need to use more advanced A/B testing tools.

4. Product Filters

Results of split testing for product filters and categories have always produced mixed results, with some eCommerce sites finding product categories unnecessary while others find them indispensable. Indeed, it’s a case-to-case basis, which is why it’s important to A/B test them, though this is by no means an easy feat. For one, you need to group each of your products in categories via the front-end, and this can be a tedious process. If you fail to do that, your testing team will be limited to what’s been grouped on the website already, thereby yielding incomplete — and therefore inaccurate — results.

5. Social Sharing Impact

Tracking social sharing is a slippery slope when it comes to A/B testing eCommerce sites. Shares via social media buttons can be tracked quite easily, but measuring its reach and network effect is still impossible as there is no technology available for split testing this yet.

The elements mentioned above may be mere drops in the ocean of eCommerce web pages, but they can create ripples of ever-increasing revenue once you optimize them properly through A/B testing. You just need the time, skills, and necessary resources to do so. In the end it will be more than worth it!