December 19, 2011
I come from a family of weather geeks. I clearly remember that we started watching The Weather Channel (TWC) every morning when I was in high school (I graduated in 1993), and we still have our favorite on air meteorologists. In fact, we are so into weather that my baby brother is working on his PhD in Meteorology at the University of Hawai’i (obviously, he is not a baby!).
So, when Chicago-based startup Weatherist popped up on my radar, I had to cover them. I was very curious about what a weather-focused website had to offer that TWC, AccuWeather, the National Weather Service, and local forecasters didn’t. After speaking with co-founder David Chung (fellow co-founder Brendan Kolber said he would be more fun to talk to), it turns out the answer is, “A lot.”
“The idea for Weatherist came from living in Chicago and dealing with volatile weather,” Chung said. “I would check one forecast and then another, and they were usually so different. I didn’t know which forecast to rely on or which one was more accurate.“
So he started tracking local and national forecasts by putting everything into a spreadsheet and comparing the forecasted weather to the actual weather. Some people were more accurate – and there were huge variances too. He realized he had something here – by culling the data and doing some very basic analysis, he had uncovered the most probable forecast.
Weatherist looks at weather from a consumer perspective, not an expert perspective, which is their competitive advantage. As an independent source for forecasts, they don’t care who is more accurate, they just want to find the most accurate forecast. By tracking the weather over a given time period, they score each source for 4 metrics – high temperature, low temperature, chance of precipitation, and amount of precipitation.
Then they basically “cut and paste” from each source to create a “daily weather fantasy team.” Every day that team changes based on past accuracy. To make sure their model worked, they created forecasts for the time period of November 1 – 27 and found their fantasy team approach was more accurate than anyone else.
“We also allow registered users to add their forecasts on our website,” Chung told me. “This lets weather enthusiasts compete with the pros, and if their forecasts are accurate enough, they can even show up on the leader board alongside TWC and local TV stations. It really levels the playing fielda and adds a layer of accountability for the big guys.”
As great as this concept is, I really wanted to know how they would make money beyond advertising. Chung replied that as their database grows, the forecast data could be valuable to transportation companies who need to plan for staffing, shipping, and fulfillment and retail analysts would like to know how traffic to stores was/will be affected. The plan right now is to create a product or API that will be attractive to those groups.
Right now, Weatherist is testing the waters in 10 cities (Atlanta, Austin, Boston, Chicago, Denver, Des Moines, Minneapolis, Orlando, San Jose, and Washington, DC) before scaling up to cover every zipcode in every city. They will also be rolling out native iPhone and Android apps as well.
Did you like this article?
Get more delivered to your inbox just like it!