FieldnoteStudio — Case study
01 · Overview
Fresh produce laid out on a table

Harvestly: bringing the corner shop online

How we helped 40+ independent grocers win back their neighbourhoods with a calmer, fresher way to order same-day food.

Mobile app 12 weeks Product design E-commerce
About the client

A marketplace for the shops you already trust

Harvestly connects neighbourhood grocers with local shoppers for same-day delivery. Founded in 2019, it grew by word of mouth — but its ordering experience never kept pace with its ambition.

Aisles of a well-stocked grocery store
The product challenge

Growth was stalling at checkout

One in five orders was abandoned at scheduling, grocers couldn't keep stock listings current, and shoppers hesitated to buy produce they couldn't see or pick. Repeat rates flatlined.

Team members discussing around a table
Our contribution

End-to-end, research-led design

Over 12 weeks we ran discovery, shaped the product strategy, redesigned the core ordering flows, and delivered a tested high-fidelity prototype ready for engineering hand-off.

Designer working on a laptop at a desk
Project outcomes

Within one quarter of launch, repeat orders climbed 38%, checkout completion reached 92%, and missed delivery windows dropped by two-thirds.

The full story — numbers, client feedback and what we learned — is waiting at the end.

Fresh, colourful meal bowls prepared with delivered produce

Explore the case study

Three short chapters — pick where you'd like to start.

Team standing around sticky notes during a discovery workshop
Chapter 02 · Challenge & Research

Understanding the problem before designing

Three weeks of fieldwork with shoppers and grocers gave us a shared, evidence-based picture of why Harvestly was losing orders — and where the real opportunity lived.

The business problem

Unreliable windows, invisible shelves

Delivery slots were promised against guesswork, so 21% of orders arrived late or missed entirely. Grocers updated stock by hand, and customers paid the price in trust.

Printed charts and reports analysed on a desk
User needs

What shoppers told us they wanted

Freshness they could verify, delivery windows they controlled, and a human on the other end when something went wrong. Grocers wanted stock updates that took seconds, not evenings.

Researcher interviewing a participant at a table
Research methods

Tap a method to see how it shaped the design.

Nine conversations with grocer owners and dispatch leads mapped the operational reality behind every order — and surfaced the stock-sync problem early.

9 interviews · 2 cities

Twenty-four shoppers logged two weeks of food runs. The "freshness doubt" pattern — hesitation before buying produce sight-unseen — appeared in 80% of diaries.

24 participants · 14 days

Five moderated rounds, 32 participants, testing everything from lo-fi paper flows to the final prototype. Each round tightened scheduling and checkout before a line of code was written.

5 rounds · 32 participants
Key insights
  1. Trust hinges on control. Shoppers want to pick a delivery window per basket — not accept an account-wide default.
  2. Freshness must be visible. Provenance photos and harvest dates reassured shoppers more than any rating system.
  3. The basket is social. Shared household lists already drove 41% of multi-item orders — the old design ignored them entirely.
Design team collaborating around laptops and sketches

From insight to interface

A six-week design sprint turned research insights into a tested, produce-first ordering experience.

Product strategy

Reframe: a freshness platform, not a delivery app

We repositioned Harvestly around verified freshness — harvest dates, provenance photos and real courier capacity — so every design decision served trust first, convenience second.

Strategists reviewing work on a laptop together
User flows

Four flows carried the whole experience. Tap a step to see what happens inside it.

Discover. Search by dish, shop or season — results ranked by harvest date, with provenance photos on every card.
Design iterations

Three fidelity levels, each validated with shoppers before moving on. Tap to enlarge.

Prototypes
Team walking through a clickable prototype together
High-fidelity prototype — 5 rounds of moderated testing, 34 fixes shipped into the final spec.
Final experience

Calm, produce-first, built for repeat runs

Harvest badges and provenance photos front the browse experience; scheduling is a per-basket, two-tap decision. The result is an app that feels like the shops it serves.

Chapter 04 · Results & Takeaways

What changed after launch

Numbers from the first quarter live, measured against the pre-redesign baseline.

0% Repeat orders
0% Checkout done
0% Fewer missed
Measurable outcomes
  • Repeat orders up 38% quarter-over-quarter, led by shared-basket households.
  • Checkout completion hit 92%, up from 61% before the redesign.
  • Missed windows down 64% thanks to capacity-backed scheduling.
  • NPS climbed from 31 to 54, with “freshest option” the top-mentioned reason.
Harvestly team celebrating results together
Client feedback
Fieldnote didn't just redesign our app — they redesigned how our grocers run their week. Harvestly finally feels like the neighbourhood it serves.
Portrait of Maya Chen
Maya Chen Chief Operating Officer, Harvestly
Project learnings
  • Constraints beat features. Showing only real, capacity-backed slots converted far better than infinite choice.
  • Design for the operator too. A 10-second stock update for grocers mattered more to freshness than any shopper-facing feature.
  • Test the doubt, not the UI. Probing why shoppers hesitated — not just where they clicked — unlocked the whole direction.
Our impact

A template for local commerce

The freshness-first pattern we built with Harvestly now anchors our playbook for local-marketplace clients — and the system is expanding to three new cities this year.

Studio team planning the next engagement around a laptop
```