When Failure Means Success

Syngenta a global seeds and chemical corporation, one of the four biggest in the world, had recently purchased a scrappy start up firm called Farmshots. Farmshots was a software that utilized satellite imagery to show farmers a heatmap of their field so that they could monitor the growth of their crops. Farmers could compare one image of their fields to another to see any deadspots in their fields, which could indicate a pest or growth issue with their crops. Syngenta had recently rolled in this scrappy team with the rest of their product offering, so it was up to me as a Product Designer to help align their product with the rest of Syngenta's applications. I worked on several projects with the Farmshots team but the one that really stands out to me was for a project that never got fully developed, and saved Syngenta money because of it.

Background Information

The Business Objective

A project manager heard from an intern that heard from a farmer that there was a need for monitoring specific areas of their field. The idea was called the monitor feature. The idea was that if an area in a farmer's field was experiencing low growth or none at all, an application could leverage satellite imagery to survey the area for several weeks and push notifications to the farmer about any changes or progress. This in theory would allow the farmer to receive up to date information about their field and allow them to make decisions like ordering pesticides to spray that particular area. In theory this would save the Farmer time and money from not having to spray as much product. Syngenta, being a low maturity UX organization, would typically take this objective and develop an application but with Farmshots, I had more say in the team and convinced the team to allow me to conduct user research to validate the idea first.

The Current Application

A bevy of my projects with Farmshots included iterating on the design systems and UI of the Farmshots application. Below are the screens of Farmshots before the monitor area feature

The current state of the Farmshots App

By the time the monitor feature was requested, I worked on several iterations for Farmshots' UI. Being that Syngenta was a low UX maturity organization, the designs I made were mostly untested and based on what I thought were effective.

The idea was that a farmer would open the Farmshots application on their phone and choose a farm to view — farmers often had many farms to look after. Once in a farm, they would get brought to the latest satellite image of their farm. To toggle between older images they would interact with the date dropdown located above the map. The real meat of the application was to look at the heat map of a field to view the health of the farmers crops. The farmer could do this by tapping on an outline of their field on the map. From there they could toggle the heatmap or add a note to the field.

The Monitor Feature

The monitor feature was a subfeature of the notes. If a user saw an issue on one of their fields, they would write a note and then within the notes section, add the monitor feature to get alerts on a specific area of the field. Short on time, I went and forgone wireframes in favor of using existing design components to update an older version of the UI.

The monitor feature

To set up a monitor area the user would pin a note to a field and then go into the note to add the monitor feature. The idea was that a farmer would likely want to add a note when they wanted to monitor an area so it made sense to pair them together, on that assumption. Looking back, I'm not sure it was a great assumption, as it may have been better to just have a separate button for monitoring, although the screen was already nearing its limit of buttons. Due to our findings from user research, I never had to make that kind of UI decision.

Initial Research

As mentioned before, Syngenta is relatively new to the concept of UX, therefore it was extremely rare that a project ever included user research upfront. However, even after designing the monitor feature, I felt pretty strongly about getting user feedback before developing such a feature, as I felt we needed to get more user feedback rather than listen to this one farmer's request.

Because of my rapport with the team and their satisfaction with my projects, the team decided to let me go ahead with the user research and set up interviews with agronomists and salesmen from Syngenta.

Who We Interviewed





Who Is Using This Application

During our interviews we learned that there were primarily three types of people that had knowledge of Farmers behaviors.

The Syngenta salesperson was our insider on working with Farmers. They were knowledgeable about Farmer's attitudes and painpoints.

Agronomists are the pesticide and crop experts. They are scientists that understand the land better than anyone. Farmers lean on these experts for their advice on dealing with pests.

Notably missing are Farmers — we did not interview them as it was in the middle of planting season.

What Are Our Users Saying?

After conducting several interviews it started to become crystal clear that there were all sorts of problems with our initial assumptions of the monitor feature. Most notably

When dealing with pests, Farmers need to spray the entire field rather than just a small area. The reason is that by the time a signifier of pests shows up on a heatmap, the pests have likely spread out to areas beyone what shows up on the satellite imagery.

The other insight we learned was that from a labor standpoint, it was easier, faster, and cheaper to send out a pesticide specialist to spray down an entire field rather than give them very specific instruction of spraying a particular area.

These two insights were enough to sink the monitor feature on sight.

Project Debrief

After presenting my insights with the rest of the team, we collectively decided to scrap the monitor feature. While one could look at the project as a waste of time, I believe the insights we learned was a success in saving Syngenta several thousands of dollars in development costs. It was also a big win in showing the value of upfront user research.

Moving on from the Monitor Feature

With the monitor feature dead, I spent the next few weeks redesigning Farmshots. One of the things I changed was could easily navigate between different fields. The image dates were tricky, as they were the same field but taken at a different time, and therefore were different maps in the eyes of a farmer. I settled on a dropdown at the top so that a farmer could easily switch between different dates to compare. These designs were validated through user testing, a first for Syngenta.

Above: The most recent iteration of Farmshots