Polyculture Research is a full-stack data science agency focused on building trustworthy analytics for companies with a commitment to sustainability. We’re based in Petaluma, California, and work with clients around the world.

The team

Devon Sampson Wills
Analytics Engineering | Causal Inference | Environmental and Sustainability Indicators

Justin Fung
Data Engineering | Full-stack Data Science | ML Engineering

Bryan Haven
Marketing Analytics | Customer Acquisition Strategy

Chelsea Sampson Wills
Business Strategy | Administration

Why polyculture?

Polyculture is the practice of growing more than one crop in the same space. Think the famous three sisters, maize, beans, and squash, of the Milpa; or traditional rice paddies that produce rice but also freshwater shrimp and wild green vegetables. After a three-year academic research project on polycultures in Southern Mexico, Devon and Chelsea were fascinated with diverse gardens and farms and all the skill, creativity and knowledge it takes to run them. We started Polyculture Research in 2015 to do applied research in that kind of agricultural ecosystem. He ran international projects for the UN Food and Agriculture Organization and the National Science Foundation. We started specializing in big synthesis projects, combining data from many sites and contexts to find new insights. That work became increasingly quantitative and statistical, and before we knew it, Polyculture Research was doing data science.

When the pandemic limited our ability to travel to field sites, Polyculture started working with software startups. A lot was new about that world, but our background in ecology served us well: finding signals in complex and noisy data, pairing urgent questions with methods that could actually answer those questions, thinking systemically while staying grounded in evidence – those were transferable skills.

Polyculture helped more than a dozen startups, venture capital firms, and nonprofit organizations design and implement their data and analytics systems. For some it involved a new data stack from scratch, and for others it was an overhaul of an existing data stack that had become unwieldy, hard to maintain, and had slowly eroded trust in data. Our job was to fix the data, and we were good at it. While each organization and their data were unique, we found that the path to building trustworthy data systems was similar at each.

We put polyculture on ice for a few years while Devon led the data science team at VSCO, a photography startup company. Now we’re re-launching Polyculture with a return to our environmental roots. We see a need to measure both business metrics and sustainability metrics, and we think we can use many of the methods and tools we used with software startups to do both, better than any other solution out there.

We know businesses that have a real commitment to sustainability have to hold more than one goal in mind all the time. They grow their revenue and their customer base, and they grow their ability to care for the environment, their workers, and their communities. Sometimes those goals are in tension, and many times they support each other. It’s a lot like growing more than one thing, like a polyculture.

Let’s Talk

We’re always interested in talking with sustainability-focused companies about their data challenges.

Get in Touch