CB Insights launches Trends, a new prediction tool

Trends uses machine learning to show what new technologies are trending based on media coverage

As a journalist, part of my job is to try to identify the upcoming, and ongoing, trends in technology. I do that by looking at what the analysts are talking about, as well as what the big companies are doing, what they're investing in and what kinds of companies they're buying. 

They can be hard to spot sometimes, even for someone who is looking for them, so for the average person, who doesn't spend their days scouring the tech blogs, it would be even more for difficult. Yet, for many industries, understanding and identifying these trends could have an impact on their bottom line. 

That's why CB Insights, a private company database that provides real-time information on companies, has launched a new feature to make that process easier. Called Trends, it applies machine learning to a massive amount of media articles, in order to enable a data-driven, real-time method to discover, predict,  plot the rise and fall of expectations and excitement when it comes to emerging technologies. 

"We have collected a massive amount of media news articles, blog posts, and long-tail trade publication content, and we realized over time that the media is a collective indicator of technology trends," Anand Sanwal, Co-founder and CEO of CB Insights, told me in an interview.

"The big question we get from clients, who are typically big enterprises, is 'Where is tech going?' Historically we would rely on pundits and industry analysts, using mostly anecdote, but we learned that when it came to building trends, there is an arc to follow. There is the early discovery, which is more niche publications, and passionate early adopter bloggers. Over time things gain widesread coverage."

CB Insights can use Trends to follow the stories as they go from those small blogs to the national news outlets. That''s why the company will "try not to take an editorial view on which publications are on the index."

"Our view is that the more fringe bloggers and vertical publications are often where the most interesting tech discussions are happening," Sanwal said."We want to vacuum up as much information as possible."

For CB Insight clients, predicting tech trends can gives viability to their corporate strategy, tells them what to focused on, what product to announce. what partnerships, investments, acquisitions to make.

A product group that is launching a wearable, for example, can see what trends are gaining momentum over the course of six months, and what to highlight as they launch. This is more of a short-term use basis. On the other hand, a strategy and innovation team can use Trends for a long-term project, using it to see what's going to be next on the horizon, and what the trends may be three or five years out.

"Folks want to understand if this is actually something material, something significant, and Trends is a good pulse check on that," said Sanwal.

From the app product management perspective, for example, people were talking a few years ago Bitcoin, but that has shifted toward blockchain. Now banks who may be building on top of blockchain protocol, can use Trends to able to see how that is continuing to shift. Or, in another case, vehicle insurance company would be able to see the autonomous car trend, and build a business around it accordingly. 

Even innovative and cutting edge companies like Google or Facebook might able to use Trends, to show their markets what they are working on and sell them on the idea.

Industries need to see what is gaining traction and what competitors were doing and now they can see news trends in context of other indicators, like company’s business momentum or industry funding, as well as get details on companies, investors and acquirers.

"We really think of Trends as early warning system. In my perfect world, someone would be consulting it right at the moment they are thinking about part of a concept or market they want to enter. This gives them the ability to flag things they should know about," Sanwal told me. 

"We also have the ability to take a media signal and do other interesting things with it, like combine it with sentiment analysis. So, here's the sentiment, and here's how it changed from positive to negative. We can parse the media and give them to trends they should be aware of and get them get deep on a trend."