Transforming a consumer grocery app into an enterprise bulk ordering platform for professional kitchens
How might we expand into the B2B market while leveraging existing infrastructure, without cannibalizing the consumer product?
How might we enable head chefs to efficiently order bulk quantities for meal planning, when current tools are optimized for individual consumer purchases?
Picnic was a successful mobile-first grocery delivery app serving individual consumers. The company wanted to explore the B2B opportunityspecifically targeting restaurants, catering businesses, and institutional kitchens that needed to order ingredients in bulk.
This was a redesign and platform expansion project. The existing consumer app had strong infrastructure (supplier network, logistics, payment systems) but was designed around consumer shopping patterns: browse, discover, add-to-cart, checkout. Professional chefs needed something fundamentally different: recipe-driven ordering, bulk quantities, recurring orders, and budget management.
I conducted interviews and contextual inquiries with head chefs, kitchen managers, and catering coordinators to understand their current ordering processes:
Through research, I identified the key jobs-to-be-done for professional users:
Start with menu, automatically generate ingredient list
Convert recipe quantities to bulk ordering units easily
See cost per dish and per order in real-time
Save and repeat previous orders with minimal effort
I explored multiple approaches to the core ordering experience, testing with target users:
Key insight from testing: Chefs didn't want a complex meal planning toolthey already had systems for that. They needed a fast way to turn their existing meal plans into orders. This led to the "Quick List" approach rather than built-in meal planning.
One constraint was maintaining the existing infrastructure while serving different user needs. I designed a responsive experience that worked across devices but optimized for the tablet-in-kitchen use case:
Chefs input their ingredient list (typed or pasted from their existing planning tools), and the system automatically matches items, suggests bulk quantities, and adds everything to cart in seconds. This respects existing workflows rather than forcing new tools.
Automatically converts recipe quantities (e.g., "200g butter") to optimal bulk units (e.g., "1kg package") while showing cost per serving. Chefs can adjust quantities with bulk-optimized controls rather than incrementing individual items.
One-tap reordering for recurring menu items. Chefs save common orders (e.g., "Weekend Brunch Menu") and reorder with a single action, making minor adjustments only when needed. This addresses the 80% recurring orders pattern.
Persistent budget indicator shows running total, cost per dish, and budget status throughout the ordering flow. Critical for professional kitchens where margins are thin and every order must stay within budget.
Schedule deliveries around kitchen prep times with supplier-specific windows. Different from consumer delivery, this respects commercial kitchen hours and loading dock access.
This was a design exploration and prototype project. The following metrics are based on usability testing and pilot program feedback.
reduction in time to complete an order compared to existing ordering methods
average usability score from head chefs in testing
of test participants said they'd prefer this to their current ordering tool
"This actually understands how we work. I'm not learning a new systemI'm just getting my existing process done faster."Head Chef, 120-seat restaurant
My initial designs included elaborate meal planning features. User testing revealed chefs already had systems they likedthey needed ordering, not planning. The best product design sometimes means doing less and integrating better with existing workflows.
I learned that "professional tools" doesn't mean "more complex." Professional users actually need simpler, faster interfaces because they're task-focused and time-constrained. Feature bloat was the enemy of adoption.
Working within existing infrastructure forced creative problem-solving. Rather than rebuild from scratch, I had to find elegant ways to serve fundamentally different user needs with the same backend systems. This constraint led to better, more realistic solutions.
If I could revisit this project, I'd spend more time understanding the full stakeholder ecosystemnot just head chefs, but also suppliers, delivery teams, and kitchen staff. Some design decisions assumed smooth handoffs that might not exist in reality. I'd also explore the invoice reconciliation flow more deeply, as this turned out to be more important to buyers than I initially anticipated.