Vinsight gives grape and almond growers a high-tech crystal ball

Farmers know how hard it is to predict yield accurately. High-value crops like grapes or almonds prove particularly challenging. Farmers expect, and settle for, twenty to forty percent uncertainty with these crops. Vinsight founder and CEO, Megan Nunes, who grew up in the farming business, found that unacceptable. almost a decade-long career in the satellite industry, she turned her tech knowledge back home. Her Redwood City, Calif.-based startup now offers forecasting software and data analytics to farmers who are growing almost anything that’s not corn, wheat or soy. While Vinsight did not have permission to name its early customers, the CEO says, it is already working with one of the world’s largest wine making businesses and the world’s second largest producer of almonds.

Rather than installing its own proprietary sensors or weather stations in the filed, or scanning farmland from on high, Vinsight amasses all the data that it can from growers, government agencies and other sources. It analyzes the data in aggregate, and comes to understand when optimum yield is likely or not, based on a wide variety of correlations.

For example, even if you experience a perfect balance of sunny days and rainy ones, if windspeed is high enough, grapes in your vineyard will shatter and won’t be salable. Or, if your decidedly more durable almond trees bloom too quickly, they won’t yield the maximum amount of nuts they could. In an ideal weather year, that could be due to bee flight hours and a lack of pollination.

Nunes said Vinsight has hit a 10% error rate, which is a step in a positive direction, and three-times better than industry standard. Here’s how they make it happen, she said. “What we do is constantly analyze and take in data on a daily basis, including from remote sensors, weather stations and satellites, to identify what is happening on a farm as it correlates with crop performance or yield. We also look at data trends on a ten to twenty year historical series. That means we’ve taken a lot of data out of old Excel files and run it through our system.”

Vinsight will compete for a share of the farmer’s wallet against scores of other agtech providers, including startups like Fruition Sciences or Arable, and others who promise to help farmers accurately predict what will happen in their fields. However, Nunes sees more room for collaboration than anything else. “We give growers insight that improves the bottom line. If you have a great data set, we want to work with you. Now, processors are using us for market research too, asking thigns like will almonds be high or low this year and where? Who do we buy from and how do we hedge our costs?”

Farmers pay Vinsight per acre analyzed, while processors and government offices pay per seat for use of the market data and software. The startup, which launched this week and was part of Y Combinator’s latest batch, is planning to expand beyond California to work with farms in Australia this year, which will help the company keep tweaking its predictive analytics all year, thanks to the juxtaposed growing seasons.