Overview

Bublibu was an internal product initiative at Paton.dev, born during the COVID-19 lockdown period in Bandung, Indonesia. It was a food delivery marketplace with one radical design principle:no owned drivers[. Instead, Bublibu was built to plug into a newly emerging ecosystem of homebrew local couriers independent logistics operators who appeared almost overnight when the pandemic paralyzed established supply chains.

The project explored a deceptively simple hypothesis:

Can a food delivery marketplace be structurally leaner if it distributes the capital expenditure of owning a driver fleet?

My Role on the Team

Bublibu was built by a team of six:

A community manager
An UI/UX designer
A graphic designer
Two developers
A QA engineer and
A CEO / Business owner and as a venture from Paton.dev.

We are working autonomusly in Bandung while the dev was working from Bali. I was the product researcher, and sat upstream of all of them under supervision of the CEO. I report my work directly to the CEO's.

here is what I owned end to end during the product Discovery phase:

Actor mapping.

I defined who the three users were (home seller, local courier, buyer), what each needed, and where the friction lived for each.

Conducting Seller& Buyer Interview.

I work with my community managers to arrange meeting with all actors, his job is to find viable early adopters while i was preparing the interview plan. at eraly stages we engage with several high qualify named culinary vendors in Bandung such as RumkayDago, Bia Kombucha, Ngombe Jamu, Bebek Setan, Dapur sedap nikmat, and many more.

Hypothesis formulation. The "no owned fleet" architectural decision originated from my research into the unit economics of food delivery platforms. I brought that framing to the team and to leadership as a product thesis, not just an observation.

Actor mapping. I defined who the three users were (home seller, local courier, buyer), what each needed, and where the friction lived for each. This became the foundation that the UI/UX designer used to structure flows, and the developers used to architect the system.

Research-to-design handoff. I worked directly with the UI/UX designer to translate research findings into design constraints — for example, the decision to make seller onboarding extremely lightweight came directly from what I learned about how home sellers operated. Research didn't end when design began.

Product framing for leadership. Because this was an internal venture, I also played a role in communicating the product direction to the company. The community manager's work in reaching home sellers and local courier operators was grounded in the user profiles I had developed.

What I didn't do is equally worth stating. I didn't art-direct the visual design — that was the UI/UX and graphic designer's domain. I didn't manage the sprint cadence or QA process. My role was to make sure the team was building the right thing, not to manage how they built it.

In a small team on an undefined project, the researcher's job is to replace chaos with clarity. That's what I tried to do here.

Context: The Invisible Marketplace

During the COVID-19 lockdown, two parallel economies silently emerged in Bandung:

1. Home-based food sellers Mothers and home cooks who had always sold food — through WhatsApp broadcasts, personal networks, and word of mouth — continued doing so. They weren't new entrepreneurs. They were existing micro-sellers operating without any digital platform infrastructure.

2. Homebrew courier operators With millions unemployed and movement restricted, a new category of logistics actor emerged: small-scale, locally operated courier businesses with no prior infrastructure. Within months, Bandung alone had at least 10 identifiable new local courier operators, including:

Ahsan Express
Paketin
Tiketux
PHK Express
BDOJEK

These operators filled a genuine gap. They were moving packages — mostly food — between homes, at a time when the major platforms either couldn't reach or weren't servicing their neighborhoods.

The core observation: Significant transaction volume was flowing through an ecosystem that had no formal marketplace, no discoverability layer, and no digital infrastructure. We called this the invisible marketplace.

Problem Statement

Existing food delivery giants — GoFood, GrabFood — operated with high marketplace fees, partially driven by the massive capital expenditure of maintaining proprietary driver fleets. This made it structurally difficult for small home-based sellers, with thin margins and low average order values, to participate profitably.

Meanwhile, a new wave of local couriers existed and was scaling organically — but remained invisible, informal, and unintegrated.

The gap: No platform existed that could serve the home seller segment while leveraging this emerging courier infrastructure — without rebuilding what the giants had already built.

Our Hypothesis

If we can distribute the capex of nurturing our own driver fleet, we have a structural chance to make marketplace fees meaningfully leaner.

This led to Bublibu's core architectural decision: build zero driver infrastructure. Instead, integrate with local couriers as logistics partners, effectively treating them as a federated fleet.

This was not just a cost decision — it was a bet on the trajectory of the local courier ecosystem. Our assumption was that these operators would mature, scale, and eventually expose APIs or integration-ready systems that Bublibu could consume.

Research Process

Phase 1 — Field Observation

Rather than starting with existing data (none existed for this market), research began with direct observation. We documented the emergence of local courier operators in Bandung by name, reach, and operating model. We identified at least 10 active operators during the lockdown period.

This was ethnographic product research — understanding a market that had no prior literature, no comparable case studies, and no playbook.

Phase 2 — User Identification

We identified three distinct user actors:

ActorPainWhat they neededHome seller (ibu rumahan)No digital storefront, limited reach, relied on WhatsAppA lightweight way to publish their food to a wider audienceLocal courierNo aggregation platform, no steady demand pipelineConsistent order volume and discoverabilityBuyerWanted local, home-cooked food that platforms didn't offerAccess to informal sellers with reliable last-mile delivery

Phase 3 — Competitive Analysis

We mapped the structural cost model of major food delivery platforms — specifically the relationship between driver fleet ownership, operational overhead, and marketplace fee percentage. This informed our core thesis that eliminating driver fleet capex could compress the fee structure.

Phase 4 — Concept Validation

We prototyped the seller onboarding flow and delivery request flow, with the courier integration designed as an open slot — ready to plug into partner APIs as they became available.

4. Underserved segments often look invisible

Home sellers broadcasting on WhatsApp had been operating for years. They weren't waiting for a platform — they had built workarounds. Good product research uncovers workarounds, because workarounds reveal real demand that formal solutions have failed to serve.

5. Hypothesis-first product development

Rather than building features, we built around a thesis. This created internal alignment and helped the team make design decisions coherently — every feature was evaluated against whether it supported or diluted the core hypothesis.

Reflection

Bublibu didn't become a commercial product. But the exercise produced something more durable: a framework for thinking about marketplace design in constrained, informal economies. The invisible marketplace we observed in Bandung in 2020 was not unique — versions of it exist in every city where formal platforms have left gaps.

The project demonstrated that you don't need proprietary infrastructure to build a marketplace. You need a clear hypothesis, a real user who is underserved, and the discipline to design for them specifically.

Case study written by Bob Hartanto — Product Researcher & Product Designer at Paton.dev Published at bobhartanto.com/casestudies/bublibu-agnostic-food-delivery