The adage "everyone's a critic" has never been more true than it is today. Long gone are the days of needing to rely solely on surveys to figure out how consumers feel about your product or business. Now -- whether it's through social media, online reviews, customer service emails, call centers or chat bots -- there is an overflow of customer feedback available for every type of business, content, or product out there.

This is a problem technology helped to create, where we are asked to give everything a review, but thankfully, technology is now helping businesses sift through and understand all of that data floating around. And for good reason.

According to a report last year by IDC, organizations that analyze all relevant data and deliver actionable information will achieve an extra $430 billion in productivity gains over their less analytically-oriented peers by 2020. Additionally, Forrester's Artificial Intelligence predictions for 2017 estimated that companies that are truly "insights-driven businesses" will steal $1.2 trillion per annum from their less-informed peers by 2020.

So how are these "insights-driven businesses" making sense of this ever-increasing stream of consumer data? To get a better idea, I reached out Dr. Catherine Havasi, CEO of Luminoso, a startup spun out of the MIT Media Lab that makes AI-based analytics software for companies like Sprint and Staples. According to Havasi, the last few years have brought significant advances in what we are able to discover from all those tweets, reviews, and online conversations. Specifically, she pointed to innovations in an area of AI called Natural Language Understanding, or NLU.

You're probably wondering what NLU is (it sounds like a strange chair you'd find in an IKEA catalog). The best description I've found is that it teaches computers how to comprehend human language as we naturally speak and type it. (In other words, trying to make sense of our "natural" language -- in all its vagueness, colloquialisms, and jargon.)

But what, exactly, can NLU help businesses learn from customers that wasn't possible before? Havasi broke it down into three categories.

Understanding your customers' "whole voice"

Positive. Negative. Neutral. What does that tell you? The answer is probably not that much. Traditional methods of analyzing language have focused heavily on being able to identify positive and negative "sentiment" in data -- i.e., thumbs up or thumbs down. But even people who are happy with a product might be critical of one aspect of it. Subtleties like this are lost with older methods of interpreting language.

NLU, on the other hand, is able to look at the full spectrum of human emotion, leading to specific, actionable nuances in data. For example, the terms frustrated and disappointed might show up in the same "negative" bucket with sentiment analysis, but NLU would pick up on the distinction between these discrete emotions. People may be frustrated with a feature that isn't working properly, but they could also be disappointed in a product that lacks a feature in the first place. Two very different insights with significantly different solutions.

Breaking the language barrier

As businesses become more global, companies increasingly need to take into account feedback from people all over the world who are using their products, not just those who speak English. But it turns out teaching a computer to comprehend a bunch of reviews in Spanish or Mandarin is much more complicated than just running a quick translation algorithm -- as anyone who has seen the movie "Lost in Translation" knows. The computer needs to be able to comprehend and analyze that language in its native tongue, otherwise much of the meaning is lost.

"Having a computer understand not just one language, but all languages at once -- that's the holy grail of natural language understanding," said Havasi. Over the last few years, she says, NLU systems with these kinds of multi-language capabilities have been progressing at a rapid pace.

New solutions for new modes of communication

Customers today want to talk to brands and companies in the same way they talk to people. This is clear in all the @-replies to companies on Twitter, the massive rise in bots, and products like Amazon's Alexa, which we can have an entire conversation with. As a result, while we have more customer feedback than ever before, that feedback is coming in a format that is new and much more conversational, where customers expect their voices to be heard and responded to quickly; if not in real time then in near-real time. It's very difficult to do that at scale without some form of NLU.


By 2019, IDC predicts that the U.S. market for big data and business analytics solutions will reach more than $98 billion. Additionally, other research from IDC shows that, for companies, improving customer relationships and becoming more data-focused, are quickly becoming the top two business goals across industries.

As Havasi aptly states, "Companies focused on customer satisfaction perform better, and you can't really know how to satisfy your customers unless you are taking their feedback and acting on it. The challenge is, up until recent advances in artificial intelligence solutions, it was really hard to scale that using old methodologies."