Till 2015, Skymet had developed the reputation of a Delphic Oracle when it came to monsoon forecasts. A perky tech start-up headed by a former journalist, it outdid the government's Indian Meteorological Department (IMD) in the accuracy of its forecast for several consecutive years, predictably making the latter the butt of many an ironic joke.
This monsoon proved a rude reality check. Skymet predicted rainfall at 102 per cent of the long period average; IMD put it at 93 per cent, suggesting below normal rainfall for a second successive year. In June, IMD stuck out its neck and revised its forecast to 88 per cent, even as Skymet confidently stuck to its prediction.
Early monsoon rainfall suggested that, yet again, Skymet's prediction would be right. But when the monsoon finally retreated in September, it was IMD that was proved right. With June-September rainfall at 86 per cent of the LPA, India suffered a drought for the second successive year.
In a country where millions of livelihoods depend on rain-fed agriculture, the accuracy of monsoon forecast is more than just an esoteric competition between weather forecasting services. Accurate forecasts are vital to farmers' sowing and harvesting decisions, central and state government's budgets hinge on them, so do insurance payouts and even the marketing and production plans of a range of consumer goods companies (since a good monsoon can be expected to boost rural demand).
Which is why Skymet Founder and CEO Jatin Singh, 38, took the trouble to publicly explain why Skymet got it wrong. In an article in The Indian Express on October 14, he admitted that Skymet "overapplied and overengineered" its forecast this year. "I had to do it," he explains ruefully in his Noida office, because monsoon forecasting is, ultimately, the "technology demonstrator" for his core weather monitoring business.
Not that this has put a damper on his ambitions for his unlisted 12-year-old company. Fresh from funding of $4.5 million last year from an arm of the Daily Mail Group and Omnivore Capital, an early-stage venture capital fund that took 33 per cent in Skymet in 2011, Singh is bustling with plans that include a raft of investments in technology, buy-outs, partnerships and honing his company's climate science expertise.
Currently, over 3,000 automatic weather stations and a field team of 100 capture and collate data inputs for the agri-business and weather forecast solutions, which account for 78 per cent of Skymet's Rs 30-crore revenues. Over the past year, Singh has invested in 15 drones to supplement this by capturing crop yields and related data, claiming to be the country's single largest user of these flying remote-controlled robots, with the aim of offering such services to state governments.
So far, he has done projects with four states - Maharashtra, Gujarat, Rajasthan and Uttar Pradesh - and claims this is the fastest-growing segment of the business. In fact, Singh says he is keen for the central government to invest in his company.
As part of this broad plan, Singh says he also plans to invest in small aircraft like the Cessna to expand into aerial photography. He's also exploring micro-satellite technology -"That way, we'll be able to count each leaf!"- and to this end he's looking at buying or co-investing in companies that offer such services (most of these are based in the US).
A new opening
Within the weather measurement rubric, Singh also sees opportunities in climate change. Hail, for instance, has become more frequent, and not just across the foothills of the Himalayas but also further south in mainland India. Although hail cannot be forecast, Singh says it can be measured - again, data critical for the crop insurance business. Currently, India has no clear hail measurement system, says Singh. He is working on a plan with an American partner - he won't say who - to offer this service.
But for Singh, who has built Skymet on the back of a family business supplying IMD - "My father was one of those Nehru Place computer vendors" -, the bigger piece of the action is in the website skymetweather.com. He wants to make it part of an integrated platform that is both global and local.
The overseas footprint is focused on SAARC countries. For instance, Skymet is working with the International Finance Corporation, the private sector lending arm of the World Bank, on a crop insurance project in Bangladesh and is trying to work with the World Bank and USAID on projects to set up meteorological infrastructure.
In India, he is also focusing on "micro-climating", or what he calls "squeezing the data" to offer customised data products that will cover 600,000 villages in 600 districts (right now, Skymet's data products work at the tehsil or sub-district level).
It's a huge business opportunity, he says, straddling the financial markets (stock, commodities, insurance, banking), government, consultancies and even individual consumers like farmers. Right now, the bulk of skymetweather.com's revenues come from advertising. From the next monsoon, Singh hopes to include a pay model offering customised services. Skymetweather.com currently accounts for about 20 per cent of revenues but Singh sees it growing so fast that it could bring in half the revenues in a couple of years.
Climate change and the growing unpredictability of weather are likely to ensure a growing demand for the kind of services Skymet offers. But high entry barriers in the form of prohibitively high investments in supercomputing and technology ensure that there are few commercial competitors - there is Kolkata-based Express Weather, for instance, at the domestic level, and websites like AccuWeather at the global level. But, says Ajit Tyagi, former director general of IMD, "There is scope for a lot of value addition in this industry."
If there is a concern in this nascent business, says Tyagi, it is the scope for manipulation. "After all, you are dealing with commodity and insurance businesses in which thousands of crores are at stake." It is a hazard that magnifies because of the lack of a regulatory mechanism, he adds.
Then of course, there are the risks involved in the complexities of forecasting and, as Tyagi points out, "the monsoon is even more tricky to forecast." In that sense, Skymet learned a salutary lesson this year. But Singh draws confidence from the fact that the bulk of his technical team is drawn from scientists who worked with the Indian Air Force's meteorological service with its formidable reputation. As for next year's monsoon forecast, Singh says he's 70 per cent sure Skymet will get it right.
This monsoon proved a rude reality check. Skymet predicted rainfall at 102 per cent of the long period average; IMD put it at 93 per cent, suggesting below normal rainfall for a second successive year. In June, IMD stuck out its neck and revised its forecast to 88 per cent, even as Skymet confidently stuck to its prediction.
Early monsoon rainfall suggested that, yet again, Skymet's prediction would be right. But when the monsoon finally retreated in September, it was IMD that was proved right. With June-September rainfall at 86 per cent of the LPA, India suffered a drought for the second successive year.
In a country where millions of livelihoods depend on rain-fed agriculture, the accuracy of monsoon forecast is more than just an esoteric competition between weather forecasting services. Accurate forecasts are vital to farmers' sowing and harvesting decisions, central and state government's budgets hinge on them, so do insurance payouts and even the marketing and production plans of a range of consumer goods companies (since a good monsoon can be expected to boost rural demand).
Not that this has put a damper on his ambitions for his unlisted 12-year-old company. Fresh from funding of $4.5 million last year from an arm of the Daily Mail Group and Omnivore Capital, an early-stage venture capital fund that took 33 per cent in Skymet in 2011, Singh is bustling with plans that include a raft of investments in technology, buy-outs, partnerships and honing his company's climate science expertise.
Currently, over 3,000 automatic weather stations and a field team of 100 capture and collate data inputs for the agri-business and weather forecast solutions, which account for 78 per cent of Skymet's Rs 30-crore revenues. Over the past year, Singh has invested in 15 drones to supplement this by capturing crop yields and related data, claiming to be the country's single largest user of these flying remote-controlled robots, with the aim of offering such services to state governments.
As part of this broad plan, Singh says he also plans to invest in small aircraft like the Cessna to expand into aerial photography. He's also exploring micro-satellite technology -"That way, we'll be able to count each leaf!"- and to this end he's looking at buying or co-investing in companies that offer such services (most of these are based in the US).
A new opening
Within the weather measurement rubric, Singh also sees opportunities in climate change. Hail, for instance, has become more frequent, and not just across the foothills of the Himalayas but also further south in mainland India. Although hail cannot be forecast, Singh says it can be measured - again, data critical for the crop insurance business. Currently, India has no clear hail measurement system, says Singh. He is working on a plan with an American partner - he won't say who - to offer this service.
But for Singh, who has built Skymet on the back of a family business supplying IMD - "My father was one of those Nehru Place computer vendors" -, the bigger piece of the action is in the website skymetweather.com. He wants to make it part of an integrated platform that is both global and local.
The overseas footprint is focused on SAARC countries. For instance, Skymet is working with the International Finance Corporation, the private sector lending arm of the World Bank, on a crop insurance project in Bangladesh and is trying to work with the World Bank and USAID on projects to set up meteorological infrastructure.
In India, he is also focusing on "micro-climating", or what he calls "squeezing the data" to offer customised data products that will cover 600,000 villages in 600 districts (right now, Skymet's data products work at the tehsil or sub-district level).
It's a huge business opportunity, he says, straddling the financial markets (stock, commodities, insurance, banking), government, consultancies and even individual consumers like farmers. Right now, the bulk of skymetweather.com's revenues come from advertising. From the next monsoon, Singh hopes to include a pay model offering customised services. Skymetweather.com currently accounts for about 20 per cent of revenues but Singh sees it growing so fast that it could bring in half the revenues in a couple of years.
Climate change and the growing unpredictability of weather are likely to ensure a growing demand for the kind of services Skymet offers. But high entry barriers in the form of prohibitively high investments in supercomputing and technology ensure that there are few commercial competitors - there is Kolkata-based Express Weather, for instance, at the domestic level, and websites like AccuWeather at the global level. But, says Ajit Tyagi, former director general of IMD, "There is scope for a lot of value addition in this industry."
If there is a concern in this nascent business, says Tyagi, it is the scope for manipulation. "After all, you are dealing with commodity and insurance businesses in which thousands of crores are at stake." It is a hazard that magnifies because of the lack of a regulatory mechanism, he adds.
Then of course, there are the risks involved in the complexities of forecasting and, as Tyagi points out, "the monsoon is even more tricky to forecast." In that sense, Skymet learned a salutary lesson this year. But Singh draws confidence from the fact that the bulk of his technical team is drawn from scientists who worked with the Indian Air Force's meteorological service with its formidable reputation. As for next year's monsoon forecast, Singh says he's 70 per cent sure Skymet will get it right.