Looking at both sides of the marketplace
On the customer side, we focused on navigation and search, paying attention to how quickly someone could move from a broad grocery need to a specific product. The shopper review covered the tools used to locate items, handle availability, and complete an order, including cases where a convenient customer action added work during fulfilment.
We followed complete tasks from beginning to end, including searching for an item, choosing a replacement, and resolving an unavailable product during fulfilment. Along those paths, information disappeared between stages and the interface sometimes asked the customer or shopper to repeat a decision the platform had already recorded.
From critique to proposal
We translated the recommendations into an interactive product concept covering the customer and shopper flows. Building the transitions between screens forced us to define what each action changed in the order, how the other side would see it, and what happened when an item was unavailable. We used one order journey in the final presentation so the judges could follow the changes from product search through fulfilment.
In the final presentation, we walked through one order from search to fulfilment using the prototype, then presented the operational assumptions and financial analysis behind each recommendation with the customer and shopper context still visible.
Making the business case
We paired the product recommendations with a financial analysis of their possible effect on retention and revenue. The model tied each proposed change to an assumption about repeated grocery orders, and we kept those assumptions visible in the deck so the judges could challenge the numbers directly.
We presented the final proposal and interactive prototype to Instacart executives, and our team placed in the top 10 out of 30 teams. The final deck connected each interface change to the part of the customer or shopper flow it affected and to the retention or revenue assumption used in our analysis.