Bringing transparency in agritrade for Indian farmers and traders

How do you bring a user-friendly app experience to small farmers and suppliers in rural India to trade in agricultural commodities? Here is how we made Bijak an easy to use app for an audience we don’t usually work with. 

Where does the potato on your plate come from? It has travelled hundreds of kilometres and tumbled through various mandis, trucks and sacks in its long journey to become a part of your biryani. Who are the people to make that happen and reach you at the price you paid per kilo? The farmers who work in different places across the year with seasons and soils to harvest their produce; those who load the harvest into trucks for the marketplace; and those who bring the trucks to the marketplace and bid for prices.

In the last few years, agri-technology in India has taken root to help these different players in the agricultural sector bring food to the table in cost-effective ways with strategies and solutions to work around logistical challenges.

In 2019, we helped build the android platform for Bijak, a B2B app that intervenes in this arena of agricultural trading and supply. We worked with Bijak end-to-end, from android development to engineering and product practices.Bijak is a new player in the agro-tech industry, helping buyers and sellers of agricultural commodities come together in a pan-Indian B2B platform. They help their users – traders, wholesalers, food processors – keep track of their transactions, access pricing and bring offline transactions online, and find other traders.

Bijak has raised over $12 million from top investment firms such as Sequoia Capital and others including Omnivore and Omidyar Network India. Bijak was also a part of the Sequoia Surge Cohort two. 

We took on the challenge of bringing the world of trading in agricultural commodities into the small screen of an app willingly. Our first task entailed a thorough understanding of the users of the app and their contexts. We  rigorously interviewed our clients to understand the nature of the transactions on the ground.

Our questions ranged from the basic to the specific: Who are the users? What is the nature of their business and how do they interact and can farmers be traders? Are they individuals, or groups, and how do they make payments? We learnt about the volume of credit players, factors that influence user ratings, and the cash flow cycles of traders in the present. We needed to understand how mandis (marketplaces) worked in order to account for varied scenarios. A key question had to do with whether we would be changing user behaviour from the ways they usually go about their business.

The edge case as the norm

or minimising the gap between the producer and user

The app had to be built from the ground-up and cater to the simplest device. Our end-users used low-technology smartphones with low screen resolution and small screens, and used their phones for calls or occasionally WhatsApp. Our UI/UX patterns had to keep this very specific audience in mind.

Dashboard - Hindi version
Dashboard - Hindi version
Dashboard - English version

The first hurdle to cross was to ensure that each part of the user experience was simple and clear, imitating real life as much as possible. For user sign up, we used minimal text input to avoid the use of too many characters or alphanumerics. The user had a choice of language at the very outset. There was only one input per screen so that the user had a simple process to sign up to use it. We built screens that had 1:1 fidelity.

This might seem simple and intuitive, but our users were not used to the quick vocabulary of apps that we take for granted in urban contexts. We realised that we had to work with assumption of varied app-literacy and familiarity with phones, ensuring no detail overwhelmed the user. We included a custom numpad for user input that involved numeric input, to make it clear and simple.

Phone number Input with custom numpad
Phone number Input with custom numpad
One input per screen
One input per screen

The biggest challenge of this app was to use the edge case as the norm and ensure incremental design and engineering effort was at a minimum.

We had to account for low internet connectivity and low battery as an integral feature of the app. WiFi was rare, and phones used mobile networks that could vary between 2G and 4G. Our users tended to go long hours without charging their phones, and GPS was a drain on an already low battery. This meant that the app had to work in offline mode and store all data locally, allowing the app to launch in aeroplane mode and operate anywhere and at any time.

Bringing the offline transactions online

Bringing offline transactions of agritrading online was a key piece of our work. We helped build a history of transactions by users selling and buying commodities for tracking reputation and credibility. This could then create a rating system for the disbursal of funds.

We had to find a way to build in this credibility and reputation that traders and farmers built in the real world into the app. We had to bring ad hoc payment systems into the app in one integrated place, where any party could gauge any other party.

This meant the app needed to include proof of payments in whatever form they were available: written slips, pictures of invoices, and pictures of the lorries carrying goods or written records in khatas. We built a system where users could either log the dates of payments and assurances, and attach images of real life documents where available. These transactions could be validated later by Bijak and did not disrupt the way users were accustomed to transacting.

BIjak order list
Order list
Bijak payment list
Payment List

Key business outcomes

We observed that traders began to record transactions on the app with greater frequency and in greater volumes. This reduced the need for manual data collection by the Bijak operations team, who had until then had conducted field work to capture data and details about the nature of working capital. Additionally, online transactions allowed for greater accountability and security, and any issues around reconciling balances and pending payments was made simpler.

Our app also reduced the amount of junk user data that had been seen in the previous version of the app. Bad or unusable data was now no longer a problem, especially with regard to the user’s phone number, or details around commodities and locations.

Our insights

The process of working with Bijak was invigorating and challenging. We had to build two releases every week, and maintain a sensitive user experience by constantly examining our assumptions and testing our app for various scenarios. We worked with Bijak for a duration of five and a half months.

Building for low-end devices requires product empathy and engineering effort, and we built our testing practices around this as our starting point, especially to stay in sync with the values of our clients. In a vast country of different degrees of digital literacy, building an app that would help users in their business and work was a learning opportunity that helped us gain more insights despite our varied experience. We enjoy this process of understanding by doing which always helps us bring our engineering insights around best practices to the requirements of a complex real-world scenario.