From Validation to Liftoff — How We Built GoFresh
A product perspective on how we built a B2B marketplace to aid GoFood merchant partners to purchase fresh goods and produce.
Gojek is not just an app, it’s an ecosystem. In fact, here’s an excerpt from our mission statement:
“Gojek is dedicated to creating and scaling up positive socio-economic impact on the ecosystem of users, driver-partners, business and micro-small-medium enterprise partners, as well as service providers.”
In order to further this mission, we are always looking to identify existing problems in the ecosystem, and figuring out ways to solve for those in a sustainable way.
This post unpacks one of the newest additions to the Gojek product roster. Our B2B procurement and supply service aimed at our merchant partners — GoFresh.
The precursors for disruption
Merchant partners are among the primary stakeholders in our ecosystem, of which our 400,000+ food merchants form a significant chunk. In our interactions with these merchant partners, and through independent research, we came up with some interesting findings:
- SMB and micro-merchants were spending anywhere between 40–50% of their revenue in the procurement of ingredients
- They had to spend cash for their purchases, thus leaving limited working capital for them to manage their operations
- As they procured items in small quantities, they usually bought goods at retail prices. This meant they were not in a position to enjoy discounts associated with bulk purchases
- They mostly procured from traditional markets called pasars
- They had to spend 1–2 hours daily to visit the wet market or a wholesaler to procure goods
- The quality of products in the market was not standardized
- There was a reliability issue as at times the products might not be available
There were clearly some problem statements here that we could address, and we had a broad idea of how to solve for them. Here are a few ideas which we thought of:
– Act as an intermediary for merchants and connect them with principal suppliers. This would result in them getting better quality products at better prices, with more reliable order fulfilment
– Deliver to merchant’s doorstep and provide them with the convenience that traditional markets cannot offer
– Provide them with terms of payment/financing for the orders they place with us in the form of micro loans or EMI payments
Post the ideation phase, it was time to set the wheels in motion. Of course, our daring new initiative still needed a name. We needed something that identified what we were trying to accomplish but also aligned with overall Gojek taxonomy (GoFood, GoRide, GoSend…you get the message 😁).
After some deliberation, GoFresh seemed to be the most obvious choice.
Getting Down to Business
Phase 1 — Check Business Viability
Considering we were venturing into uncharted waters with a new business line, we thought it prudent to test the business viability of this idea. Before kicking off product development, we needed to understand the challenges we needed to overcome, especially in the area of procurement, logistics, and fulfillment.
We formed a small crack team within the GoFood product group to run this pilot, and a few months into the pilot, we had a clearer idea of what we were getting into.
Our initial hypothesis had been right, the absence of major existing players in the B2B space meant the market was ready for disruption. But that in itself did not mean we could get away with sub-par execution.
We would still need robust suppliers who could guarantee the availability of high-quality fresh products, manage deliveries on time, and had cold chain capabilities. Even if we achieved this, changing the procurement behavior of merchants from traditional markets to using a digital product would not happen overnight. We would have to first have a sales-assisted ordering process before slowly moving the merchants to become self-reliant, i.e Outbound to Inbound.
Our efforts largely hinged on merchant leaders in a neighborhood making the transition and getting an experience superior to their current process. These merchant partners would then spur further migration within their community to our platform, thus driving growth.
Another advantage of running this pilot was that it helped us define the overall value proposition. Here’s what we were aiming for with GoFresh:
Phase 2 — Airdrop Product Engineering
With the business viability check done, it was time to get the product team into action. So we conducted an inception workshop with the key stakeholders from the business development and operations team to evaluate existing problems in each department based on impact, effort, and urgency.
As an outcome of this exercise we identified 3 major milestones for the product engineering team to focus on in the first 6 months:
1: Build an order and content management system for internal teams to use
2: Build an ordering platform for merchants to use, which can be integrated with our existing merchant partner application
3: Integrate with our suppliers to relay orders and manage catalogue in real time
It was important that we also consider scale while setting these milestones. If we got out estimates wrong, our systems would end up breaking, and our solution would lose the trust of our merchant partners.
Don’t believe scale can be a double-edged sword? Let me elaborate:
- If we scaled beyond x orders/day, the sales team would need to expand proportionally, which would not be viable at this early stage.
- Once the merchant community picked up that a direct ordering solution was available to them, we would be able to focus less on ramping up the sales team and more on operations (as beyond y orders/day, manual operations processes will start failing).
- Tech integration with the suppliers will remove the inefficiencies of human processes and set the trajectory for growth.
The plan was clear, but first things first, we needed to build a product team from scratch.
Phase 3 — Build
At Gojek, new use cases to be solved come up every quarter. Since we’ve done this a few times, we wanted to build a backend architecture that was easily expandable and allowed for multiple user interfaces to be built over it.
Our plan was to have a two-pronged approach:
The first one was the standard product dev plan for our internal system:
- Understand the requirements from the internal teams
- Wireframe the user flows and validate with the users
- Build the backend platform
- Build the user interface
In short, Design>Build>Test>Deploy>Repeat
The second one involved prepping for our second milestone i.e mobile app:
- Conduct research with merchants to clearly identify existing pain points, expectations, and behaviour
- Build prototypes and conduct usability tests
In short, be ready for the engineering team to start work once they are done with the first milestone.
Phase 4 — Launch
Our sales team was using an open-source solution to temporarily manage orders until the product team was ready with the order and content management system. We wanted the migration to our in-house system to go smoothly, without impacting any of the business processes.
We needed a rollout plan.
So we came up with something along these lines 👇
- Create a user/training manual and hand it over to the users
- Conduct training across various business teams
- Create sub-groups within the business teams and onboard each group in phases
- Constantly monitor for system stability and production issues while ensuring the business processes do not break, while also gradually increasing the rollout
In the course of 3 weeks, we were able to move all of the internal teams from the existing open-source solution to a new web application — no delays, no downtimes. 🙌
GoFresh is now live in Jakarta, and seeing hundreds of orders per week among merchants who have joined the platform. Fresh off the high of this first success, the team is currently going full steam ahead to launch the GoFresh mobile app in the coming months. Exciting times ahead!
Liked what you read? Sign up for our newsletter and get our stories beamed straight to your inbox! 🖖
From Validation to Liftoff — How We Built GoFresh was originally published in Gojek Product + Tech on Medium, where people are continuing the conversation by highlighting and responding to this story.