Agriculture and algorithms

November 21, 2012 Editor 0

IN INDIA, when monsoons are delayed and crops fail as a result, farmers often don’t know how to pay back the debts they have taken on to purchase seeds. More than 15,000 commit suicide every year. These fates are a shocking reminder of a global problem caused by global warming. Farming has always been a gamble, but the growing number of “unusual weather events”, as experts call them, make seeding and harvesting an even riskier business.

The Climate Corporation, a start-up based in Silicon Valley, wants to reverse the trend and reduce farmers’ financial risks—by crossing agriculture with the IT industry’s latest trend: big data. The firm is collecting all kinds of information—including on weather patterns, climate trends and soil characteristics—and analyses the data down to an individual field. These insights are then used to offer farmers tailored insurance policies against the damage from extreme weather events.

Premiums for the company’s “Total Weather Insurance” (TWI) plans depend on crop and location. On average, they cost about $30 an acre annually, some 3% of the land’s revenue. In case of extreme weather at the wrong time of the season, The Climate Corporation pays out up to $300 per acre (the TWI is designed to complement federal crop insurance programmes in America, which provide only limited cover). In contrast to existing government schemes, farmers don’t have to prove actual losses. Payouts are triggered automatically without paperwork when the firm’s data show that writing a check is justified.

“We are not predicting the weather. We estimate the likelihood of unusual weather events and their potential impact on every single field in the US within the next two years”, explains David Friedberg, the firm’s founder and chief executive. To do this, it has to be good at both analysing huge amounts of data and calculating risk. To help with the number crunching, Mr Friedberg, an astrophysicist who once worked for Google, has hired a team of “quants”, or “quantitative analysts”, from the financial industry.

Yet marketing is an equally big challenge for the six-year-old company. Farmers are least likely to be early adopters, especially when it comes to a new product that lives in a computing cloud, admits Mr Friedberg. The Climate Corporation has a website where customers can buy policies online. But it had to learn that selling its insurance to farmers in remote areas is best done through a network of agents.

As most other start-up bosses, Mr Friedberg does not reveal sales figures, saying only that “many thousands of customers have bought the product”. But venture capitalists seem to like the numbers they have seen—which makes The Climate Corporation yet another example of VCs moving away from social media start-ups and warming up to more sustainable business-to-business ventures, says Niko Waesche, the author of a book on data-driven business models. Since the company was founded in 2006 (then named “Weather Bill”), Mr Friedberg and his partner Siraj Khaliq managed to collect a total of $110m from such noted VC firms as the Google Ventures, Atomico, Founders Fund and Khosla Ventures.

So far, The Climate Corporation offers its policies only in America. But if its combination of agriculture and algorithms succeeds there, it may go global. Australia, Canada and Brazil are next on the list. Here’s hoping that insurance policies based on clever analytics will one day also protect Indian farmers against the vagaries of the weather.

Go to Source

Categories: The Economist

Tags: agriculture and algorithms, Climate Corporation, unusual weather events