Test boosts annual sales by £1.9M

• Overview

• Project scope

• Heuristic evaluation

• Big win A/B testing

• User flows

• Concept wireframes

• Responsive hi-fi designs

• Results

• Recommendation

Overview

Hush is a women's fashion and lifestyle brand in the UK. In 2023 their yearly sales reached £62 million. They planned to optimise and re-build their entire checkout process. I consulted with Hush between Jun 2023 — Oct 2023. I was very fortunate to work along side Hannah Brooker and Sara Berrada.

Project scope

I planned out the project into three key steps. The first step was a heuristic evaluation of the checkout process. This would solve two problems, how does it work and are there any immediate fixes?

Step two involved creating solutions to some of the issues found in the evaluation. I did this by designing an A/B test recommendation pack.

The final step was the redesign of the checkout. This was the most involved part of the process. This included mapping out all existing user flows, building wireframes and fully responsive designs.

Heuristic evaluation

The main objective was to assess the platform’s usability. Strengths, weaknesses, and areas for improvement were identified, scored and commented on. This insight was used to inform actionable and prioritised recommendations.

Big win A/B testing

An extensive funnel analysis provided some very insightful data. The most significant issue being identified at the login page. For every 100 customers that leave the cart to the checkout, 66 drop off. This was a significant spike in numbers compared to the remainder of the user journey.

Despite having this data, it was unclear as to why so many customers were abandoning the checkout. These were some of my findings that formed the bases of my hypotheses for testing:

• The “red banner” is a prominent distraction to the user. This could increase drop off rates

• The “default header” is available. This could increase drop off rates

• “Login” is the default option. “Guest checkout” content is hard to find and hidden beneath a unconventional “button style tab”

• The “default footer” is available. This could increase drop off rates

• The overwhelming majority of users give up trying to login if they can’t remember their username or password

• Customers may not know all the benefits of “guest checkout” vs “logging in”

• You can only view one option at a time. This guarantees that a portion of users will have their preferred option hidden

• A user is forced to provide an “email” before being able to continue into the main checkout area

• There’s no option for customers to create an account at this initial stage of the checkout

This lead to three recommendations for testing:

Test one

The fastest and most cost effective test to implement. Defaulting to the “guest checkout” rather than “login”.

Test two

This time presenting both options to customer, leading with “guest checkout”. A more involved test to implement, however tackling more known issues.

Test three

The ideal solution, tackling all issues identified but requiring a full redesign.

Providing three options gave us lots of flexibility. Being unsure of budget and resource, we could be confident we had a workable option. We initially implemented test one which showed positive results. This gave us enough credible evidence to implement test three.

We set this A/B test up as a priority and ran it in the background whilst I stared on the checkout redesign.

User flows

This process started with mapping out the existing user flows. This helped me to understand in detail how the checkout was currently designed and built. I then mapped out each possible user flow, totalling up required fields and interactive elements.

Significant gains can be made reducing the number of form fields. Therefore, knowing running totals of all users flows is crucial.

Concept wireframes

Mapping out all users flows gave my the core building blocks to re-imagine the checkout. This time my goal was to use fewer moving part and re-order the content into more logical groupings. This enabled me to reduce the amount of questions by about 50%.

This formed the bases of my initial concepts, mocked-up as sketches. This approach allowed for quick idea generation. Despite being rough sketches, these were enough to share the initial concepts quickly.

After validating the proof of concept I moved onto the next step. Crafting new user interface components was necessary for these designs. These were built in Figma where I mapped out the initial UI designs. These prototypes enabled us to test the rigours of the user flows and micro-interactions.

Responsive hi-fi designs

After carefully designs the core user journey I was able to complete the full checkout user flows. These development ready specs were then handed over to the development team for building and testing.

Up until this point every idea, sketch and prototype was mobile first. The last few steps were building out tablet and desktop breakpoints.

Results

We initially saw positive numbers over the first few days of the test. Unfortunately these number then declined, with the final results being inconclusive. Based on the strength of test one and the initial hypothesis we decided to re-built the test. This time we used a developer to ensure we had a robust build and UI.

After a four weeks we saw:

Increased projected annual sales ↑ £1.9M

Increased users moving to checkout ↑ 2.1%

Reduced exit rate from log in page ↓ 8.4%

This created a significant uplift in projected annual sales of 3%. This will now earn Hush an additional £1.9M in sales every year.

∙ ∙ ∙

I had the pleasure of working with Andreas on a some UX Optimisation projects during our time together at Hush, he added so much value to the projects offering his broad experience and expertise. Andreas was very reliable with a great attention to detail, and patience around lengthy discussions/back and forths on design decisions and sign offs.

Andreas was able to use his diverse experience and technical expertise to really think about the end user, which reflects in his intuitive designs. He demonstrated strong communication skills throughout our time working together, listening to feedback and incorporating it to refine his work.
Further to that, Andreas was just great to chat to in general and glean insights from. He has a naturally positive outlook and offered up a new way of looking at things.

I highly recommend Andreas to any team seeking a dedicated and talented UX designer.

Sara Berrada
CRO Specialist

∙ ∙ ∙

You can view all the topics featured in this case study at hush-uk.com