Digitise contract farming in the developing world

Can we develop a user-focused UX for a mobile app and web dashboard that can benefit one million African farmers?

My Role

Product Manager.  User Research, Product Design, UX design, UI design, Interaction, Branding, Prototyping & Testing, Product Strategy, Pitching

Client

Multiple

Sector

Agriculture, Farming, Sustainability, Food Security

Project Date

2017 to present

Background

The Challenge

When my boss at the Creative Industries KTN left London to move to Malawi, I didn’t expect to hear from him again so it was a pleasant surprise when he reached out to me almost 2 years later and invited me to come to Malawi to help out on a project all about modernising the age-old contract farming industry.

In contract farming, a company will typically engage with thousands of individual local farmers and enter into a contract with each one promising to buy their produce. They would often give the farmers cash or seed loans, give them training to help them increase their output and want to have evidence that they are helping to improve the livelihood of the farmer over time. Over the past few years, the requirement for traceability (evidence of the provenance of the produce all the way back to origin) has increased as has the requirement for fair trade, organic certification, sustainability. In the developing world, there are still large numbers of farmers who have low literacy levels and do not have a national ID card to verify who they are or prove ownership of their land. Crucially, many do not have internet access or a smartphone.

Therefore there are still many companies with teams of ‘Extension Workers’ travelling to visit farmers to collate data they need on that farmer’s ID, her farm, her production, her loan repayments, her families welfare and more, using paper and pen or on a phone and sent into an office where a human will translate it all into a database. The companies have real problems knowing who is registered, who will honour their loan agreement and predicting what the upcoming harvest will look like.

A large organisation in Malawi was in this situation. They had a team of 30 extension workers.  They asked the innovation consultancy I worked with to look at the solution for them. That’s where I come in! With John as lead developer and myself as lead designer, we got to work!

Market Research

Competitive Analysis

Back in 2017 when we started, it was very much an emerging market and there was very little established competition. My desk research found that the few solutions out there at the time fell into 3 groups, and further user research gave me an insight into how these solutions were perceived:

Competitor Group 1

Off the shelf farm management software

It was clunky and rigid. It wasn’t a good fit to how we do things”.

The cost was far too high for our budget”

Competitor Group 2

Free / low cost field data collection tools

Basically, a survey tool as mobile app or in a mobile web browser –

We need someone with experience in this tool to make it work for us. We don’t have the know-how to use it properly”

Competitor Group 3

Bespoke software development

Hire an agency or employ in-house development and IT support team –

Recruiting people out here with the skills we need is really difficult and expensive. And then he moves on he takes the knowledge with him, we have to recruit all over again.”

Research

User Profiles

Having lived in London nearly all my life, I had to get to grips with what life was like for subsistence farmers in rural areas in the developing world and the organisations working with them!

I spent a week ‘in the field’ conducting interviews with all the key stakeholders and user groups. I defined the main user profiles and based them on real interviewees to better empathize with the key stakeholders and prioritize goals according to their needs.

User profile A: Conscience
User Profile B: Victas
User Profile D: Caitlin
Early stage ideation

Lo-fi Wireframes

We received some funding prior to the first commercial client to do the initial research and create an MVP. Where we developed concepts such as

  • Modelling the real world by creating data models for each key entityFor example: famer, field, village etc
  • Using a semantic database to create links between the entitiesi.e. Farmer A is linked to Field B as OwnerOf from date X to Y
  • Cropping profilesA data model for how and what is produced each season - linked to the farmer and the field
  • Visit cyclesA UI for the field worker to schedule and keep track of the farms he needs to visit

It wasn’t until we had our first commission from FSU that allowed me to do further research and testing with real Extension Team Workers out in the field and develop concepts like:

  • Chunking: Replace long forms with smaller onesI observed Extension Workers sit with each farmer in the hot sun for huge amounts of time filling in long forms. This made both parties tired and prone to giving answers that they hoped would speed up the process rather than give the right answers. I suggested ‘Chunking’. Information and tasks grouped into familiar, manageable units are more easily processed and more rewarding. By splitting up a complex task into a sequence of chunks, you can effectively simplify the task. Each chunk represents a separate mental space, easier to deal with alone than as a whole
  • Work plans (instead of visit Cycles)User configurable Work Plans set up on the Mobile App to manage routine field worker activities. Enabling the managers to set tasks and monitor progress across these tasks, as well as allowing field workers to see their progress in real time. Visit Cycles proved to be impractical in real life as most Extension Workers changed plans daily based on the weather and community happenings (funerals, religious holidays etc).
  • One question per screen We found this simplified the use of the app and made it more pleasurable. It also worked better on smaller screen sizes.
  • Mobile app UI designed for high visibility in bright sunlight and minimum battery drain
UI Design

High Fidelity Mockups

After testing the wireframes via 121 interviews with sample users, I made some adjustments and began to create the high-fidelity wireframes/mockups.

Below are some of the web analytics dashboard screens. There were several technical constraints and pre-conditions that needed to be taken into account which I had to bear in mind during the design of the web UI. I had worked with the same development team on previous dashboards for other products and time was limited so I knew we had to use similar frameworks & toolkits to build out this one rapidly (bootstrap, angular js, highcharts etc).

Similarly, with the mobile app UI, we had limited development resources and iOS devices were virtually non-existent in the environment we were designing for! Therefore, I stuck very closely to the Android Material Design system to speed up development time and provide a UI that users could use even if it was their first time using an android device.
The below designs are for tablet in landscape mode.

Transition from Product Design to Product Ownership

Product Management

Within 18 months we had gone from just an idea to a product being used by multiple organisations. My role expanded to a Product Manager as we looked to onboard more clients in different countries with nuanced needs. In August 2018, after listening to users on another visit to Malawi to see our largest client and also speaking with potential clients, I wrote a document that informed our product roadmap and client acquisition strategy for the next few years:

Roadmap plan for a client in Ghana (JIRA screengrab)
Iterate & Roll Out

Expansion

For the next 3 years, as Product Manager my role involved:

  • Creating of product brand including logo, visual identity, marketing materials and website
  • Leading the rollout of a customised version for each organisation. Including sales demonstrations, UI modifications, user training and, in many cases, digital transformation consultancy.
  • Requirements gathering and user testing workshops in-person in various African locations and remotely (due to Covid-19)
  • On-going management of product roadmap - gathering user feedback and usability data, translating that into continual improvements.
  • Leading the development team. Prioritization and management of development and maintenance using JIRA
Iterate & Roll Out

New Features

As Product Manager and UX design lead, I’ve researched, designed and implemented several new features in to the product over the years

  • Perimeter mapping
  • Honey – batching, beehive location, bucket tracking
  • Traceability
  • In-app Yield estimate calculations
Key Highlights

Results

  • Successfully launched the product for first client in Malawi with a team of 30 extension workers
  • Our product enabled them to scale up to a team of 120 covering the whole country within 18 months
  • Used our learnings to commercialise the product suitable for a large variety of organisations
  • Currently used by multiple organisations in 8 countries across Africa, directly impacting over 500,000 smallholder farmers

Get in touch 🤙

Let's create something great together!