Focusing on conversion rate can lead to lower sales and profits. Revenue per visitor is often a more telling metric.

Johann Van Tonder, chief operating officer, AWA digital

Johann Van Tonder, chief operating officer, AWA digital

Why do 95% of your website visitors leave without making a purchase? And what you can do about it? Based on our business experience, and our success rate with our clients such as Canon, Interflora, Avis and a growing list of SME’s, we offer the following thoughts.

Conducting A/B testing, where two groups of users see different variations to determine which one is better, is synonymous with CRO (Conversion Rate Optimization). You’re in good company if you do this. Leading companies including Amazon, Google, LinkedIn and Booking.com actively pursue strong experimentation cultures, achieving incremental growth from constant testing.

When Thomas Edison was asked about his many failed experiments, he reportedly replied: “I have not failed. I’ve just found 10,000 ways that won’t work.” Think of a/b-tests as your own scientific experiments, designed to determine what matters to your users by deliberately shaping behavior and measuring the outcome on business KPIs. Even if the outcome of the test is negative, you will learn something about user behavior, which can be leveraged.

Driving more conversions comes down to an obsession with people, and not visits.

Failure to grasp laws of statistics that govern this experimentation can however lead to the wrong conclusions being drawn and potentially hurt sales. In fact, two thirds of reported A/B test results are said to be bogus. There is an over-reliance in the industry on statistical significance. Albeit an important indicator, it should be seen in the context of other factors such as statistical power and confidence intervals.

advertisement

A related debate is how long A/B tests should run for. It is risky to observe common rules of thumb such as “x number of conversions per variation”. One good way to solve it, is to use one of the many test duration calculators available online.

The phrase “Conversion Rate Optimization” is perhaps a misnomer. It would suggest that the primary objective is always to turn more website visitors into paying customers. In fact, it is not advisable to obsess over conversion rates. In isolation, conversion rate is fairly meaningless. It can be increased at the expense of overall revenue and, ultimately, profit. In e-commerce, a more appropriate metric is often RPV (Revenue Per Visitor), a composite of conversion rate and AOV (Average Order Value).

Your testing program will only be as effective as the quality of ideas fed into the system. You need to generate ideas to inform an evidence-led approach. Running with hunches has its place, but it’s not a foundation for producing consistent results. Shooting from the hip presents an opportunity cost as a more valid idea could have taken the place of a random test that may or may not deliver.

This intelligence gathering involves looking at hard numbers as well as engaging with your visitors directly. For example, you might use Google Analytics to map out customer journeys to see points where visitors are dropping off the site. But in order to do something about it, avoiding a hit-and-miss shotgun approach; you also want to consider why it’s happening. For that, you could pose well-targeted onsite questions or conduct usability testing.

advertisement

Panel-based usability testing is widely used, but we recommend a slightly different methodology called Remote Moderated Usability Testing, whereby users are recruited live from your site to observe their actual session in real time. Often, the most unexpected – and lucrative – insights have been generated in this way.

Merchandising Analytics is an area that doesn’t receive much coverage, but ultimately your website converts because users are convinced by your merchandise and how it’s presented. With the help of commonly available tools, merchandise analytics can help to get the right product in front of the right person at the right time. It can even inform your product mix, the introduction of new products or indeed the retirement of products that adversely effect loyalty.

Because optimization is concerned with changing behavior, it’s important to be familiar with relevant aspects of consumer psychology and the user’s convoluted journey to the purchase decision. Google’s Avinash Kaushik remarked: “Driving more conversions comes down to an obsession with people, and not visits. Johann and Dan’s book explores this subtle but critical difference in depth. Understand people, their behavior, their feedback, their testing choices – and win bigger revenues.”

Based in the United Kingdom, AWA digital is a conversion rate optimization agency. Johann Van Tonder is chief operating officer at AWA digital, and co-author, with Dan Croxen-John, of E-COMMERCE WEBSITE OPTIMIZATION:  Why 95% Of Your Website Visitors Don’t Buy, And What You Can Do About It.

advertisement
Favorite