Kickstarting a data-driven CRO program for Metals4U increased sales by more than 34%

When we run an experimentation program for our clients, our aim is always to start driving tangible results from the very first day.

From our initial analytics review, we found that a large proportion of Metals4U’s users were exiting their website on the Product Detail Pages (PDP). Digging deeper, our usability tests showed that at this stage of the funnel visitors were often looking for information about delivery times, which they couldn’t find.

It turned out that in their eagerness to find the price, many mobile users were scrolling straight past the delivery messaging. Even for those who did read the messaging, there was no information on the page about delivery speed.

We hypothesized that if users were able to see the delivery speed for each product, and if delivery information was presented in an easily digestible format, they would be significantly more likely to purchase.

For our first experiment, therefore, we chose to test this hypothesis by prominently positioning the delivery information and speed near the top of the PDP.

The result? These changes increased the conversion rate by 6.9%, adding more than £580k in annualised revenue.

And not only did this experiment produce a very healthy conversion rate uplift, but it also validated our user research. Clearly, delivery information was extremely important to Metals4U’s customers – this was a learning we were able to take with us and apply to future experiments to great effect.

In fact, the increase in demand caused by this run of experiments put so much strain on Metals4U’s warehouse that we were asked to pause the program until they had a chance to adapt to this new level of demand.

‘Trusting’  Conversion to scale experimentation

At Conversion, we’ve developed a scalable approach to experimentation, meaning that the more we test, the better our results.

One of our early experiments for Metals4U was focused around trust.

Our analytics research showed that many new users were failing to make the final purchase once they’d arrived at the checkout process.

User research suggested that this might have been caused by a lack of trust: purchase sizes were often quite large, but many new users were unfamiliar with the Metals4U brand. Combined, these factors seemed to be deterring new users from completing their purchases.

We were confident that if we could find a way to make these users trust the Metals4U brand, we could increase the likelihood of them completing their purchases.

As it turned out, many of the usual hallmarks of website trustworthiness were not present on Metals4U’s website, which meant that there were plenty of opportunities for us to start building trust.

For example, we added payment labels (for MasterCard, Visa, American Express, and Paypal) and included payment security messaging on the PDP and checkout journey.

Not only did this experiment produce a 4.8% increase in the conversion rate, but it also validated our hypothesis that trust was an important factor in their users’ purchasing decisions.

This was a solid result in and of itself, but our process is about more than creating one-off conversion rate uplifts – it’s about learning from our experiments, building on our progress, and creating as large an overall uplift for the program as possible.

As such, we chose to take the learning from this experiment and apply it to a variety of different areas across the site, even going so far as to build the entire homepage proposition around the idea of trust.