Customer data becomes marketing insights

With more than 50 offices in 27 different countries, dunnhumby is a global leader in customer data science, providing insights to the world’s largest retailers and CPG brands, including Tesco, Coca-Cola, Procter & Gamble and L’Oréal.

In the UK specifically, dunnhumby analyzes purchase data from top retailers such as Tesco to provide clients with powerful insights to help them better understand customer needs, product performance and category trends.

Led by Karyn Mukerjee, Director of Data Innovation, dunnhumby came to us with their vision for the HuYu app (available on iOS and Android), which invites shoppers to scan grocery receipts from selected retailers and answer surveys in return for rewards, including shopping vouchers.

All data would be anonymized and aggregated before analysis. With these insights, retailers and could better understand their customers and improve their products and services.

Dunnhumby looked to leverage our delivery expertise to rapidly get the product to market, which would let them start building a user base. In practice, this meant using a collaborative problem-solving approach and building trust in order to make key technical decisions to enable the team to move quickly.

Our design and engineering excellence complemented dunnhumby’s own specialism in customer data science, laying the foundation for the perfect partnership. A trusted relationship grew as the joint teams collaborated across design, engineering and delivery.

We had a clear vision for the HuYu app and focused on the need to balance user experience with the needs of retailers and CPGs — for example, which features would be good for users and generate data which could create the most powerful insights? … We built a strong joint team and were able to deliver a great product in a very short time. The project has built a solid foundation and we have exciting plans for the future.

— Karyn Mukerjee, Director, Data Innovation and Partnerships at dunnhumby

While creating a consumer app was a new venture for dunnhumby, the team was set up to embrace this venture with a lean startup model. By identifying the business outcomes and embracing an experimental mode of working, our joint team was able to learn and adapt quickly to the market and minimize waste.

We focused on these key ways of working to facilitate rapid learning and iterations:

•  Establishing clear business and user outcomes and KPIs from the start

•  Taking an outcome-driven approach to prioritizing features and analytics

• Choosing technologies that enabled fast delivery and change

• Focusing on reducing cycle time (the time from starting work to delivering value)

Dunnhumby was eager to release the first iteration of HuYu quickly in order to start gathering data, learn what worked for users and understand the potential for clients.

As speed became a critical success factor for the project, we recommended and implemented a series of cutting-edge technologies to deliver the first release of HuYu in just under 12 weeks. Those technologies included Flutter, serverless architecture, Google Cloud Vision and Firebase.

Our joint team of product owners, engineers and designers were uniquely cross-functional and collaborative. Bolstered by the guidance and expertise of Lead Engineer Swav Kulinski, a trusted leader and ambassador of Flutter, the team delivered:

• Value at pace through unblocking problems and upskilling quickly in a cross-functional manner. Designers and testers picked up coding, and engineers learned new languages like Dart and delivered full-stack stories.

• Rapid capability growth, as the team learned and adapted quickly. That was all facilitated by the cutting-edge tech stack.

• Quick problem-solving through tight feedback loops and the decision to adopt serverless architecture, removing the provisioning or managing of physical servers.

•  An exceptionally smooth handover to dunnhumby at the completion of the project due to the collaboration and extensive upskilling within the team.

• High levels of user satisfaction. According to an in-app survey of 2,000 users, 87% found the app easy to use while 80% would recommend HuYu to a friend.  

• A successful first release in 12 weeks allowing users to scan receipts, fill surveys and earn points. A second release enabled rewards redemption and customer spending insights.

• The entire architecture was built using Flutter and serverless architecture based on Firebase and GCP, enabling us to deliver both iOS and Android apps with one code.

• Use of native components, making it easier to develop new features and keep the app up to date with new releases.

• Quick and easy app testing facilitated by the Flutter and Firebase relationship. This resulted in tight feedback loops and immediate deployment of changes.