If you’re a college student in Buenos Aires or Chennai, you may have come across an unorthodox way of making extra money. Using your Android phone, an American corporation will pay you to stop by the supermarket on the way home; snap a picture of how much bread or tomatoes costs that day; and submit the price of those commodities into an elaborate data system. California-based Premise, as they're called, uses this information as fodder for an unusual business model: getting inflation and commodity price data before governments do, sourced from regular people on the street.
Using thousands of college students and other part-time workers, Premise gathers raw item prices from retailers and street markets worldwide. The information Premise’s workers collect is used to help develop live inflation indexes and food security data for clients including hedge funds and government agencies. David Soloff, Premise’s CEO and the former head of analytics firm Metamarkets, told Co.Labs that the company’s goal is to apply crowdsourcing and digital information-gathering technology to the still "post-World War II" methods of generating inflation data.
"We see people making really big important decisions like printing more money," Soloff says. "But they're basing those decisions on information collected through a data supply chain that arose in the years following the Second World War. Provisioning of actual data points in these old supply chains leads decision makers to make bad decisions." In other words, Premise believes they can offer decision makers more accurate global financial information more quickly.
While Premise’s strategy sounds ambitious, enough venture has been put behind the company since their 2013 founding to make it worth paying attention to. This past March, Soloff’s firm raised a $11 million Series B round lead by Social+Capital Partnership. Other investors in Premise include Google Ventures, Harrison Metal, and Andreessen Horowitz.
According to Premise, the company is currently collecting economic data in the United States, Brazil, Argentina, India, China, Japan, and Australia. Crowdsourced part-time employees, mainly college students, download a proprietary Android app which gives them "tasks" to complete in their city in exchange for financial compensation. Once the employee photographs a price tag or a for-sale sign, that information then goes through the company’s servers for QA and evaluation by supervisors in the United States. The crowdsourced data provides one half of Premise’s data stream; the company also uses price information scraped from tens of thousands of e-commerce websites.This information is then used to generate real-time indexes for Premise’s customers.
Right now, Premise’s efforts are relatively modest. The $11 million they raised in March is intended toward expansion into additional international markets and bulking up their presence in markets where they are already operating—as of October 2013, they only had 22 part-time employees covering all of Mumbai. According to a Premise publicist, the company currently has 1250 part-time employees worldwide with that number expected to expand to 40,000 by the end of 2014.
Premise is currently offering their indexes to corporations and financial service providers, to government agencies, and to marketing organizations. Of these, Soloff feels government agencies have the biggest potential for licensing Premise’s indexes. "Government banks own huge data capture buckets, a huge trove of data expertise. They have a lot of wisdom but don’t necessarily have fast-moving data pipelines. We see them as a natural customer set, and we want to learn how to be a data provider for some of these entities," says Soloff.
As a secondary revenue stream, Premise is also considering licensing their crowdsourcing model and workflow pipeline as a market research tool. According to Soloff, Premise’s platform can be used to keep tabs on corporate competitors’ product placement, and on where counterfeit products show up in foreign markets.
But in the end, Premise’s biggest potential is their promise to offer more accurate economic data than governments, more quickly. In Argentina, one of their biggest target markets, the government has been accused of fudging inflation data for years. By applying a TaskRabbit or Mechanical Turk-like approach to gathering economic indicators, Premise is engaging in untested (but highly promising) methods.