🔒 Protected case study
This GoMart instant grocery case study is password-protected. Enter the password to continue.
Gojek · GoMart · H1 2021
GoMart is Gojek's instant grocery service — order from supermarkets and local stores across Indonesian cities and get hand-picked groceries delivered to your door.
Alfamart, Lotte Mart, Sayurbox, Watsons and more — with dedicated shoppers preparing orders in-store and drivers fulfilling last-mile delivery inside the Gojek super-app.
GoMart · Consumer Journey
As a Product Designer lead, I led discovery and conversion design — facilitating Studio Accelerator sprints, desk-research synthesis, and cross-functional alignment between Search & Reco, Reliability, and Brand.
Case study · GoMart
How might we help competent shoppers discover, trust, and routinely use GoMart?
Anchored on KR 2.5 — moving visit-to-book from 14% to 25% — through a Studio Accelerator sprint that turned funnel data and brand strategy into a portfolio of bets: bottom navigation as the structural unlock, plus order-tracking & chat and onboarding stories alongside.
Decorative recording of the shipped GoMart app showing bottom navigation and browse flows.
Discovery · Solo desk research, ahead of Session 1
I independently mapped the funnel end-to-end — Gojek home → GoMart home → search → merchant → cart — to give Session 1 a shared starting point: ~41% dropped before GoMart home, ~52% from search to add, and merchant pages lost ~50%.
Solution 1 · Key · Three phased releases, one tested
I designed three progressively richer releases before this went to research — each one a hypothesis about what the navbar needed to solve. Only the most complete release went into the June 2021 concept-testing study with 6 IDIs.
Release 1 · MVP
Hypothesis: promoting order history to its own tab — with a one-tap "Order again" that rebuilds the cart — will lift reorder rate for routine shoppers.
Desk research had flagged low visibility for the reorder widget and past-purchase section.
Release 2 · +Merchant
Hypothesis: giving Merchant its own destination — store grouping, branches, categories — will cut the back-and-forth between stores.
Targets the ~50% merchant-page drop-off flagged in desk research.
Release 3 · +Promo
Hypothesis: consolidating deal content into one destination will lift engagement for deal-driven shoppers — addressing the "inefficient deal browsing" and low "Deals of the Day" visibility research flagged.
→ Taken into concept testing · 6 IDIs, June 2021
One hypothesis from this release didn't hold up: a dedicated Search tab. Testing disproved it — refinements on the next slide.
Solution 1 · Key · Design & Validate · Concept testing · June 2021 · 6 IDIs
We walked participants through the full 5-tab wireframe from the previous slide — Explore, Merchant, Search, Promo, and My Orders — and tagged expectations, delights, and pains per menu.
What held up
Explore kept its name (GoFood precedent); search bar + recs added to the homepage.
Before build · Discoverability
Coachmark/tutorial became mandatory — no participant recognised the bottom nav without prompting.
Search stayed on Explore, not as a tab — matching GoFood behaviour and what participants already expected from the homepage.
Solution 1 · Key · From research to visual design
Bottom navigation wasn't a leap of faith. GoFood had proven the pattern, 6 IDIs validated Release 3, and the same five-tab IA carried straight into visual design.
Visually: wireframe placeholders became GoMart's brand system — green accents, photography, real merchant storefronts, and product-led promo cards. Structure unchanged; fidelity upgraded.
Drag to compare wireframe and visual design
Highlight · Solution 2 · Trust · Effective Communication
Trust · mid-order
A parallel Studio bet on trust after booking. During an active order, consumers chat with both shopper and driver — from the Gojek homepage track widget, Chat tab, or an in-flow modal. Born from service blueprint gaps: no customer–shopper chat, verbal consent only.
