Case study: Transforming a customer personalisation journey

Creating an inclusive, optimised personalisation experience.

I’ve omitted confidential information and key metrics in this case study due to privacy agreements. All information presented is my own and does not reflect the views of Wonderbly.

TL;DR: We built an optimised, customer-centric avatar builder that increased personalisation options by 12,500% and led to double-digit revenue growth.

Company background

Wonderbly is an online personalised children’s book company that recently won ‘Children’s Publisher of the Year’. Since 2013, their bestselling, inspirational books have delighted over six million children around the world. Customers visit their site to buy a book for a child in their life which they can personalise- see an example here.

My role

I joined Wonderbly in 2020 as a product manager and lead an agile, cross-functional squad to deliver optimised, intuitive customer experiences that improve the shopping experience and drive conversion (CVR).

Case Study

This case study explores how I led a cross-functional team to launch Wonderbly’s new personalisation journey and deliver an intuitive, delightful experience that increased conversion and enabled thousands of children around the world to see themselves represented. Below, explains how we went from idea to launch based on deisgn thinking methodolgy.

The customer need

Wonderbly’s customers have a strong need to personalise their book to the extent they desire. Customers’ ‘aha’ moment is watching the child they bought a book for, see themselves and feel a positive emotion such as being inspired, feeling brave or loved etc. I’ve articulated this below in a user story.

“As someone buying a personalised book, I want to personalise the book as much as I want, so the child in my life feels a positive emotion.”

Most of Wonderbly’s books are personalised by adding a name and avatar that portrays the child. The above image shows the typical personalisation journey- a user can add a name and pick from 6 avatars in two simple steps. Whilst this journey had led to success, from customer data we knew customers weren’t satisfied and that we could better fulfil this need.

From analysing quant & qual customer data, I identified there were two key customer problems:

• Customers couldn't personalise the books to the extent they wanted.
  Often, customers couldn’t see the child in their life represented in the current selection and wanted to create a more personalised avatar.
• Customers felt unfairly excluded.
  The selection of avatars had more options for White children compared to Children of Colour. Many customers complained this was discriminatory and unjust- they felt the child in their life had been excluded.

After identifying this concerning customer problem as a high-priority project, I built a compelling business case to increase avatar diversity by adding more custom options (e.g hair, skin tone) and build a new personalisation UI. Next, I had the challenge of pitching this to the executive team and achieving buy-in.

At Wonderbly, we ‘ship small & learn fast’- we’re constantly looking at how we can build an MVP quickly, ship it, learn & iterate. Due to the technical complexity of our personlisation journey, we couldn't create a new avatar UI in a couple of sprints. To build an MVP required around 5 weeks of engineering work and 2 months of creating new artwork- I had to justify why this work was worth it. After several intense pitches, I secured buy-in and officially started leading the project!

We knew users weren't satisfied with our existing avatars, but how do we improve? What extra personalisation options should we include — glasses, skin tone, hairstyle, accessories? The human race is diverse, there were many options we could include.

To learn more about this, I worked with our publishing team and a research agency to interview diverse parents and their children. Our key aim was to find out how users wanted to be represented. This was one of my favourite parts of the project, as the discussions I had and insights discovered were fascinating.

Below are some of the key insights:

• Representation in books is extremely important, parents want their kids to see themselves and feel inspired and empowered.
• Skin tone, hair colour and hairstyle were identified as the most important features. Seeing these features in avatars made the children feel represented.
• Avatar personalisation options needed to be authentic and not based on stereotypes. This was a point many parents felt strongly about.
• We also discovered the optimal number of choices per feature- i.e. how many options of skin tones, hair colours is too much or too little.

Designing the right personalisation options

Armed with our fresh insights, I worked closely with our talented artists to develop our initial set of skin tones, hair colours and hairstyles.
Through sending a survey and conducting user interviews with diverse groups, we gathered important user feedback and iterated on the designs. Speaking to users was key to ensuring our personalisation choices were authentic and not stereotypical.

Designing a new personalisation user interface

To accommodate the extra avatar features, we needed to create a new, interactive avatar picker that customers enjoyed. This hadn’t been done before; it was a good opportunity to greatly improve the personalisation UX and delight customers. Working closely with our UI designer, over several weeks we began story mapping the key customer journey and created an initial prototype to usability test with users. From this, we improved the UI based on feedback and were finally ready to start building.

My dedicated dev and content team worked tirelessly to ship our new, exciting UI and deliver the best customer experience! There were many bumps and bugs along the journey but everything was handled with a positive, focused and empathetic approach. Once the UI was shipped, we all felt very proud of what we had achieved.

Learning through A/B testing

To learn how the new UI performed. I launched an A/B test where 50% of users would see our existing avatars and 50% of users would see our new UI.
Prior to launching the test, I defined primary and secondary metrics and sample size. My hypothesis was that CVR across our key KPIs would be higher in the improved journey.

The results

Below is a summary of what was achieved:

• Compared to the control, the new optimised journey led to a higher CVR across the key metrics and led to double-digit growth.
• Customer options choices grew by 12500% — there are now 700+ avatar combinations to choose from.
• Our new UI was rolled out to all international markets and continues to drive YoY CVR. Plus has received excellent customer feedback and led to a decrease in negative reviews and complaints.
• Due to the success of the project, senior leadership agreed to allocate further resources and budget to develop the UI & add more personalisation features.
• I also received a Wonderbly crown award for leading this project- which was awesome!

Check out what we built here.

Key takeaways

• I LOVED this project! Creating inclusive products is something I’m passionate about and want to focus on in my career. This launch challenged me to be aware of my biases, validate assumptions and learn about inclusive product development. Less than 1% of British children’s books have a main character of colour. I’m very proud to have helped more children see themselves in books & hopefully feel inspired.
• Team collaboration leads to innovation. This mammoth project couldn’t have happened without many incredible, unstoppable teams working together. Through communication together, sharing expertise, executing with empathy, we built something incredible!
• Motivating your team is essential. This was a very long project with many challenges along the way. To keep teams focused and motivated, I embraced a positive, supportive attitude, organised team breaks and listened to my teams’ concerns/ thoughts.