Caterlush - Connecticut's Local Chef Marketplace
Caterlush is an 8-month redesign of a B2C food marketplace connecting home cooks with local customers across Connecticut.
Role
Duration
Tools
Team
Project Overview
Caterlush is a Connecticut marketplace connecting home cooks and independent chefs with customers ordering homemade meals and corporate catering, built around direct trust between diners and the person cooking their food.
I was brought on as [your role] to redesign [specific scope], with the goal of [specific goal].
Just need those three brackets from you — your actual role, what part of the platform you worked on, and what you were aiming to improve — and it's ready to drop in.
Problem Statement
Caterlush's early platform struggled with 80% drop-off between browsing and completing an order. New customers found it hard to trust ordering food from someone they'd never met, and cooks found managing listings and orders more complicated than it needed to be.
40% of customers left before completing checkout, often after viewing a cook's profile
First-time visitors were 3x more likely to leave without ordering than returning customers
Cooks reported spending too much time manually tracking orders outside the platform
Mobile users had a noticeably harder time browsing and checking out than desktop users
Without visible trust signals, customers defaulted to more established food-delivery apps instead
Research & Discovery
I conducted 3 week research to understand how people search for and trust homemade food online, and how chefs manage listings and orders.
15
Users interviewed
600+
Survey Responses
5
Competitor Audits
Key Findings
Customers hesitated to order from a stranger's kitchen without clear trust signals - reviews, photos, or chef background
Chefs found it hard to manage orders, menus, and availability without a simple dashboard
Existing food marketplaces felt corporate and impersonal, which didn't match what Caterlush stood for
Ideation
I ran a short sprint with the founder to turn research findings into a clear direction — mapping the two core user types (customers and cooks) and sketching how each should move through the product.
User Flow
I mapped the core customer journey — browse cooks → view a dish or menu → order → checkout — keeping it to as few steps as possible so ordering felt as easy as any food delivery app.
Browse by cuisine or cook, with trust signals (reviews, photos) visible immediately
Simple menu view with clear pricing and availability
Streamlined checkout with saved details for repeat orders
I designed a matching flow on the cook side - list a dish, manage orders, update availability - so both sides of the marketplace felt equally considered.
Wireframing
I created wireframes for 18 key screens, starting with low-fidelity sketches and iterating to mid-fidelity grayscale layouts. Each round incorporated feedback from internal reviews and 2 rounds of guerrilla usability testing.
The final wireframes achieved a 92% task completion rate in usability tests, compared to 58% for the original design.
Visual Design
I built the visual identity around Caterlush's warm, homemade feel - moving away from the cold, corporate look of typical food-delivery apps.
Prototyping & Testing
I built a high-fidelity prototype in Figma and tested it with 4 participants across 6 rounds, covering both the customer ordering flow and the cook dashboard.
4
Test Participants
6
Testing Rounds
95%
Task Success Rate
Iterations & Refinements
Iteration 1: Simplified the cook dashboard from 3 steps to list a new dish, based on cooks getting stuck during testing
Iteration 2: Added trust signals (reviews, photos, response time) higher up in the browsing flow after customers said this was their biggest hesitation
Final Design
The final product includes a customer-facing marketplace for browsing and ordering, and a cook dashboard for managing listings, orders, and availability - both designed to feel simple, warm, and trustworthy rather than transactional.
Redesigned homepage and browse flow with trust signals (reviews, photos) visible upfront
Simplified checkout from 6 steps to 2
New cook dashboard for order and menu management
Consistent visual system - warm terracotta and cream palette - carried across every page, from ordering to onboarding a new cook
Impact
58%
NPS Improvement
25%
On boarding drop-off
10
Support Reduction
The redesign paid for itself within the first month. Our support tickets dropped by 60% and customer satisfaction has never been higher.
Lessons Learned
Designing for two very different user types (customers and cooks) on one platform meant every decision had to work for both sides, not just one
Trust is the biggest design problem in any marketplace connecting strangers - more than any single UI decision, visible trust signals drove behavior
Next Steps
Expand to new cities across Connecticut as the cook network grows
Add repeat-ordering and favorites, so returning customers can reorder in fewer taps
Introduce cook verification badges to strengthen trust signals further
