Tesco food waste
Reducing food waste across Tesco's stores
Background:
As the largest supermarket in the UK, Tesco wanted to set a precedent on how supermarkets could minimise food waste. They had already come a long way in reducing food waste in their stores, but had some ambitious targets to reduce this further.
My role:
I worked as part of a multi-disciplinary team to research problems across the existing food waste process and design iterative solutions using in-context testing and data to ensure any updates met Tesco's food waste targets.
Discovery research
Tesco had identified some areas where they were underperforming against their food waste OKRs. I worked closely with another service designer to investigate these issues across the full lifecycle of the food that goes unsold at stores. This involved ethnographic research to observe staff in stores and the waste management centres that surplus food was transported to. We also reviewed the existing technology systems which allowed Tesco to keep track of their surplus food and link up different processes, such as inventory management and reporting.
By immersing ourselves in the subject area and investigating the full spectrum of processes, touchpoints and actors for everything that goes on behind the scenes, we were able to pinpoint several problems that could be causing underperformance against OKRs. Everything we learnt was mapped to a service blueprint depicting the existing process and an assessment of what we had observed.
Photos from ethnographic research
Service blueprint
Data & low fidelity experiments
Food waste was regularly tracked at Tesco so we got hold of all the existing data in various spreadsheets and Tableau analytics to investigate any patterns and check our observations from the ethnographic research. For example, one of the issues identified related to 'contamination' were different waste food types were inadvertently mixed together. The data indicated there were repeated food types that were most often the source of contamination and also specific stores that contributed to this more than others.
We then conducted some additional quick in-context experiments to validate our assumptions. For example we set store staff a task to sort different food waste items to see how they grouped them and followed this with questions to understand their thought processes. This provided valuable insight for us to come up with ideas for how to help staff reduce food waste contamination.
Design sprints
As a multi-disciplinary team consisting of key stakeholders and subject matter experts across Operations, Technology and Tesco's Food surplus coaches we worked collaboratively involving everyone in the design process. I led design sprints for different problem areas, taking each problem through ideation, prototyping and testing in just a 5 day period.
It was challenging at first to get buy in from all stakeholders for this approach, but by involving everyone every step of the way they started to see the benefits for themselves and soon enough we were being asked to help with other business initiatives using a similar approach.
Design sprint workshop
Ideas sketches from design sprint workshop
Iterative prototyping & testing
It was important we tested everything rigorously before putting it live because at Tesco it was a big job to implement even the smallest changes to in-store processes with additional training required to all necessary store staff. We therefore tested solutions using in-context research methods to ensure we tested in realistic situations with real end users as the environment and context impacted use significantly. We used different fidelities of prototypes as necessary in the testing eg paper prototypes, Marvel prototypes on devices and functional prototypes. This was hugely beneficial as each time we tested we learned something new and would iterate the prototype to test again.
We moved at a fast pace so didn't produce formal research documentation and instead used tools like Miro boards to keep track of objectives, hypotheses and results and quick summary emails to share key takeaways from each round of testing with the wider team. We also created note templates that doubled up as low fidelity discussion guides, which were effective for conducting in-context research when we spent a lot of time cramped into back of house areas and busy shop floors!
Once the entire team were confident in the solution, we conducted pilots to trial new solutions in a subset of stores first, which allowed us to track data from actual usage at a larger scale and measure how the solution impacted performance. We used a goals, signals, metrics framework to set performance measures and selected a range of stores using existing data to create a representative sample and to truly stress test the solution. The data from the pilot was combined with learnings from observations and interviews with users who are trialled the solution before it was put live across all of Tesco stores.
Top image: Example of research objectives in Miro boards
Bottom image: Example of research note template/discussion guide
Outcome
The improvements to the food waste process overall reduced weekly food waste by 40%, which equates to saving 100 tonnes per week.
I am also proud of how we managed to embed a new collaborative approach to problem solving with the wider team at Tesco, with prototyping and testing at the heart.
