Design, Impact, People, Success Story, UX

Capturing Customer Insights; Increasing Website Conversions

Even before the coronavirus pandemic led to a drastic drop in high street retail sales, brands across Australia were increasing their online presence to meet the needs of digital consumers. This was the case for one of Transpire’s most recent customers – a well-known retailer.

Although trade in its physical stores remains solid, a strong online presence is an ongoing priority, as this area of the business continues to grow in size and significance.

To this end, it wanted to clarify the perspective of website users and gain insights into their buying journeys. Transpire’s role was to provide more confidence around digital strategy by capturing customer sentiments and insights to increase conversions and revenue.

Solution – Interviews & Exploration

The focus here was to learn about the user’s behaviour and motivations when making purchases online. Starting with an informal interview, we asked questions about the user’s most recent online purchase, followed by questions specifically about their experience shopping with the retailer in question. Finally, we asked the users to complete an end-to-end transaction to observe usability.

Of all our findings, we were most interested in the user’s checkout experience, as the key problem to solve was increasing conversions and revenue. During testing, we noticed that users would either stop or stall at the “Continue as Guest/Login” page before adding payment and delivery details.

Upon further investigation, when users were asked why they stopped or stalled, most said they were unsure whether to continue the task or wait for further instruction from the session facilitator.

We hypothesised that the user was distracted from the task by having to decide if they were a registered user (which includes thinking about when they signed up, what their password was, what email address they used) or whether they wanted to continue as a guest/register at another time.

Solution – Design by Data

In order to support our hypothesis, we needed to search for relevant data. With the help of Google Analytics and Hotjar, we could further validate our hypothesis to see whether or not it was correct.

We soon discovered that when faced with the option of signing in or checking out as a guest, 25% of users didn’t continue with their purchase – a huge dropoff. This prompted various recommendations based on UX best practice and competitor websites:

Give guest checkout priority – Change the page’s layout to emphasise guest checkout and include a smaller call-to-action (CTA) at the top for users already registered.
Divert user to guest login directly – Also include a secondary ‘already registered’ CTA so the user has the option to create an account once the purchase has been finalised.

Based on these recommendations, we began to design several different user flows where we could remove or adjust the ‘Checkout as Guest/Login’ page during checkout.

To both increase collaboration and improve outcomes, Transpire invited members of the business’s e-commerce, marketing and loyalty teams to a co-design session, also known as a design studio. This allowed for the rapid creation of ideas and solutions from the wide range of knowledge and experience brought by the session’s different participants.

After pitching, critiquing and voting on ideas, we could then move forward with the design of a solution that was technically feasible in a tight timeframe.

Because we needed to manage risk and allow the process to lead us, a decision was made to use A/B split testing to measure the impact of change. The last thing we wanted to do was make a change that adversely affected bottom line sales.

Impact

We used a popular B2C ecommerce solution to run a 14-day A/B test of the proposed design solution. It was exposed to 25% of our audience (control 75%) with the purpose of measuring conversion rate but also checkout rate i.e. the number of people who started a checkout vs. number of people who completed a checkout.

Ultimately, we wanted to get a view of the potential positive/negative impact to conversion for users who were checking out.

With evidence that our new design solution worked, the next step was to roll it out to 100% of users around the world. Additional checkout simplifications, which had already been designed, were also put forward for delivery.