Revamping the Sales Funnel for Pets at Home

I was hired by Wunderman Thompson to work as a UX Design Lead on their Pets at Home account, taking in keys areas of PLP / PDP pages, Basket and Checkout, and Subscriptions.

The client was undergoing a comprehensive rebranding exercise, and a total overhaul of their digital products was part of that undertaking.

Pets at Home had many parts to their business and bringing them together coherently was part of the challenge faced by the agency.

Too many landing pages and too many disparate brands, but there weren’t enough ways of easily navigating between their complimentary services.

The key problem I was tasked with solving was to enhance the sales funnel on the main PaH eComm site. It had too many options and CTAs, resulting in decision paralysis and high dropout rates amongst users.

There were also issues with the way some of the fulfillment options available to customers were presented in the basket and checkout. Some items would end up being thrown out of users’ baskets without their knowledge, resulting in confusion and disappointment.

A thorough assessment of all real-world delivery combinations - as well as all possible basket permutations - was required in order to apply a commonsense rationale to the way this was presented to the customer.

My team and I gathered as much information about these problems as possible via - for example - customer support anecdotes, client observations, first hand experience and remote user testing.

Triangulating our sources using a combination of test data, video recordings and questionnaires we built up a picture of how the experience looked from the user’s perspective - the what, the why and how they felt about it - and tried to define the pain points and areas where we could reduce friction.

There were user journey issues that needed sorting out - a good many of them centred around the shortcomings of the Pets at Home delivery and distribution setup.

Some of the fulfillment scenarios meant that we couldn’t provide a split basket, for example, as it was too costly to send items via DC and dropship within the same order.

This meant that the Basket and Checkout areas had to work hard (as did we) in providing acceptable solutions to these problems.

We also had to decide what to do for other unhappy path situations, e.g. in the event of an item having gone out of stock between adding to Basket and then checking out.

Automatically adjusting the Basket contents seems at first glance like the most helpful thing to do for the user, but it soon became obvious this was a disjointing experience and only confused people.

Flagging a problem and then prompting them to take action proved the best way of informing users of the issue and ensuring they knew of changes their product selections, i.e. an item hasn’t just vanished from their Basket - they have actually removed it themselves on instruction and been given a reason.

I validated this hypothesis in Maze via prototype testing with post-task questioning. The new concept scored a much higher satisfaction rating, and had a better completion rate than the existing journey.

One of the other main problems I encountered was in decluttering the interface to reduce cognitive friction. There were a lot of complications - particularly around pricing - which gave users too much thinking to do. It’s not that there weren’t good reasons for a lot of this complexity, but we pushed hard in gaining permission to decrease the amount of information required on page.

Where some additional pricing detail had to stay, for example the Subscribe and Save price option, I built some strong text hierarchy into the wireframe in order to show the user where to look first.

Overuse of CTA buttons was another issue to resolve, as this also added friction. Testing showed a high rate of stall when confronted with multiple CTAs on PDPs. So I started paring these back, for example, removing the Store Selection CTA button from PDPs and instead opting for a passive default.

Reducing to one CTA on A/B testing showed a jump of 450% in engagement (much to the delight of the client) so we proceeded in that direction.

The evolution of the design through to UI stage was a gradual process with a lot of liaising from the UXers. The application of existing digital brand guidelines meant some areas had to be revisited, with site mechanics tweaked to allow for a better all round experience.

The synergy between the UX and UI teams was thankfully very evident, and elegant solutions were found to the many issues that presented themselves.

Further hi-fidelity user testing continued, along with live A/B testing of some proposed changes to the checkout process - the client wanted to be sure of the right course of action before commiting. After all, they do turn over a billion pounds a year so a mis-step would have been costly.

The client were very happy with the end product and the company saw a notable uptick in sales after the overall design work was completed - revenue to 2023 went up by 7.9% - with our efforts being no small part of that.

The previously disparate nature of the brand’s online offering now feels like a cohesive whole, and it presents as a market leader - a far higher pedigree to its peers...