Land O’Lakes is 103 years old and for most of the dairy cooperative’s existence, it did not have a chief technology officer.
That changed in 2018 when Teddy Bekele ascended to the role, an acknowledgement that even farming needs to be more tech forward.
Companies in established markets have traditionally viewed technology as "an enabler" to accelerate growth or productivity, says Bekele, who is leading the digital transformation efforts for the cooperative's four business units. At Land O’Lakes, Bekele says, the core business mission remains the primary focus, "but I do think more and more, technology can go from being an enabler to being a driver and being at the center of what we do."
With nearly $17 billion in annual revenue, Land O’Lakes has a broad business that includes dairy products sold under the namesake brand, Purina Mills animal feeds, and WinField United’s seeds and crops. By leaning in on tech, Bekele says, the Arden Hills, Minn.-based company is helping farmers optimize their milk production, lean on smarter data to determine which crops will grow best in the field, and establish more sustainable practices.
Combining machine learning technology with Land O'Lakes' sprawling collection of agriculture research hubs has been especially eye-opening. The company, which sits at #245 on the Fortune 500, has more than 100 research plots of land all across the U.S. where it conducts testing and observes the best conditions for the various plant varietals it sells, as well as a 55,000-square-foot research hub that explores crop growth under extreme weather conditions.
Bekele shares the example of a farm in southern Minnesota, where one might think that conditions would be similar to a Land O’Lakes-owned research plot just 10 miles away. Indeed, that was once the assumption that guided the company's practices.
But when crunching the data—including soil, the climate, and topography—machine learning is delivering some valuable, counterintuitive insights. It turns out that a plot that’s 300 miles away in South Dakota might actually be more representative of the Minnesota farm than the one 10 miles down the road. Those insights are shared with retail agronomists, who know all about field crops and who can help farmers understand what would perform best on their land.
Land O’Lakes is also starting to roll out a generative AI tool, developed in partnership with Microsoft, that acts as a virtual agent and can answer questions about crop insights and soil.
For decades, agronomists used a dense, physical book to pull the relevant farming information. An effort to digitize the asset into a PDF didn’t make searching any easier. But with the generative AI tool, farmers can do things like share a picture of an unusual weed growing in their field and share it with an agronomist, who can more speedily identify the weed with the new tool and make a recommendation on how to combat it.
Bekele, who initially joined Land O'Lakes in 2013 as senior IT director and as the chief information officer of the WinField United division, has led a bigger bet on cloud, also in partnership with Microsoft. About 55% of applications runs on the Microsoft Azure cloud computing platform today, up from 30% when he became CTO. Basic applications that don’t require immediate innovation remain on prem, but “whenever time’s up, and we can do something else with it, we’ll probably rip it out,” Bekele says.
John Kell
Send thoughts or suggestions to CIO Intelligence here.
