FarmLink possesses the world’s largest source of high-quality data on farm production. The company’s data is derived from millions of yield observations collected from the country’s largest fleet of instrumented combines, and overlaid with decades of satellite imagery. This proprietary data forms the basis of the company’s unique ability to benchmark yield potential for every field.
Actionable data, not big data, is the key to increasing revenue for large and small farms, particularly when margins are tight. —Bob McClure, Chief Data Scientist for FarmLink
My team collaborated with FarmLink’s data scientists to develop a suite of products to bring precision farming to modern growers. The ask—tools that are powerful and easy to use. The promise—to maximize land use, harvest yields, and selling price, for sustainable profitability.
As senior UX designer project, my contributions included user research, information architecture, wireframing, and client workshop and presentations.
FarmLink was expanding its data analytics platform to offer a range of applications at various price points for different customers. Some of these products would result from strategic partnerships with other organizations with unique datasets.
Our entry point was FarmLink’s DTN MarketVision product. Our team had 8 months to redesign DTN and provide a roadmap for future products. DTN was functional software, but it was dense, visually outdated, and inscrutable.
You pay attention to analysts, to the USDA, to the weather. Agriculture isn’t just here on my farm. It’s global. You gotta know what’s going on in China.”
Market data is important year-round. We subscribe to multiple analyst reports. They give you their opinion, whether they are right or not is another thing.”
My research plan included stakeholder interviews, customer interviews, and market research to inform our work.
Talking to the stakeholders clarified the vision and key goals for FarmLink, and provided an understanding of what FarmLink perceived users wanted and struggled with.
To interview customers, we traveled to the fields of Iowa to meet with growers and agronomists. Going into the field—literally—and talking to potential users of FarmLink products allowed us to validate what stakeholders said about users, as well as gain additional insights about their daily tasks, decisions, digital usage behavior, and concerns.
Here are a sample of our findings across various topics:
For each chart, we first had to understand the data story it was telling. Then, we worked to enhance it. We asked: Can this data be interactive? Can we split or combine graphs to increase clarity? Can we tuck away some information and display it only if the user wants it? Can we standardize the presentation? What help can we offer the user?
We worked with the client in a systematic and iterative way, as changes to one area of data often affected to many other areas of the application.
DTN MarketVision offers farmers valuable grain marketing information in one convenient place. Farmers can rely on DTN for insights into when to sell their grain, for real-time updates on crop yield potential, and to manage risk.
For a fresh, modern aesthetic that fits in the farm environment, we created a visual design system that’s vibrant, graphic, and earthy. Visual elements include:
+ High-contrast colors support readability, even in sunlit fields
+ Page hierarchy conveys via page titles, icons, and content blocks
+ Color blocks and typography give prominence to important data
+ Contextual help and editable fields encourage interaction
After working on DTN, it became clear that FarmLink needed a powerful and flexible publishing model to keep up with content across a broad spectrum of future products. Standardized modules used across different products not only unify branding but also streamline product launches.
In order to design good software, domain knowledge is necessary—but only up to a point. Working on FarmLink, I was initially overwhelmed with the complexity of farming and of data science. I did become quite knowledgeable about key concepts and terminology, but I also realized that I did not need to fully understand farming. Nor did I need to fully understand data science.
As a designers, our expertise is in representing a user-centric perspective. We bring knowledge about what makes software usable, how to visually organize information, and what current expectations and UI conventions are to the table. We only need to absorb enough domain knowledge to be able to ask necessary questions.
I wrote a brief article about this idea in HOW Magazine.