When Apple first launched Siri for the iPhone 4S in 2011, it didn’t take long for industry insiders to realize that talking to our phones would change not only the ways we seek out information online, but also the ways in which we interact with our devices.

However, users were a bit more baffled as to how they were going to use voice-recognition on devices. Who can forget the loveable grandma trying to get Siri to understand her accent (not to mention the hoards of angry Scots)?

Fast forward seven years later, and we’re asking Alexa for cooking instructions while we’re up to our elbows in dinner preparations and talking to chatbots on Facebook Messenger about our customer service issues. So what’s different? The evolution of the natural language processing (NLP).

Put simply, NLP is function of artificial intelligence (AI) that enables computers not just to recognize, but to interpret and anticipate human language. Sounds simple, but take the classic grammar joke “Let’s eat, grandma,” and imagine how a machine that didn’t understand commas or context would interpret that. NLP helps computers not just recognize words, but learn common usage, understand accents, and predict the ways that humans speak and search.

How are we using NLP right now? – Siri, Amazon Echo, Google Home

Voice-activated devices are quickly becoming the new normal. A recent study by PWC shows that consumers ages 18-24 describe themselves as “heavy” users of NLP technology, along with 57% of those 50 and above. And those users aren’t just chatting with their devices; they’re using them to make purchasing decisions. The NPD group actually found that after purchasing an Echo, consumers actually spent more money on Amazon.

So what should brands be doing to make sure they’re searchable for Siri? Sticking to the facts, according to Apricot Law’s Tom Desmond. In the past, customers using Google were offered pages and pages of results, but voice assistants generally only offer up one or two options. For example, if you say, “Hey Siri, find a dry cleaner,” Siri offers just one name with the option to view others. To get to the top of this new, more limited search, companies need to be very specific with their online information.

“AI platforms try to respond to such queries in a human way, and they use the text of pages in search results to do so,” Desmond recently wrote for ClickZ. “Content should be optimized for conversational language with clear, grammatically correct answers to specific questions, such as who, what, where, when, and why.”

Creating landing pages with clear location, hours of operation, etc. Claiming Google’s My Business pages has become more important than ever in the age of NLP and voice search.


Chatbots get a bad rap sometimes, mostly because we know them as the smooth female voices barring our path to talking to an actual human being. But many savvy brands are using chatbots to take their customer service to the next level. For example, Starbucks has used AI technology to create a virtual barista. The company’s My Barista app uses NLP technology to allow users not only to order coffee via chatbot but also predict what customers might want to order in the future.

And this is only the beginning for using chatbots to improve customer service. According to Gartner, 85% of business will soon be done without interacting with another human.

So far, it seems like audiences are pretty on board with the chatbot revolution, with 69% of consumers reporting that they actually prefer a chatbot to an actual human. However, even though those numbers are promising, many companies are slow to experiment with chatbots, with only 7% of marketers reporting that they currently use them. But those numbers could soon be on the rise, since a further 27% indicated they were thinking of experimenting with chatbots in the near future.

Google Duplex

When Siri was first introduced, it was hard to believe that in just a few short years, AI technology would come so far that we’d be asking Alexa to help us find our lost phones. And the future seems to promise even more possibilities, starting with using NLP technology to create robots that will handle our most mundane tasks.

In May of 2018, Google introduced Duplex, a robot designed to engage in natural human conversation. This summer, the company plans to test Duplex on actual humans in order to check holiday hours for businesses, make restaurant reservations, and finally to see if this AI can pass the modern-day Turing test by making an appointment for a haircut.

What’s next?

Google remains at the forefront of AI and NLP technology, and has recently announced plans to begin testing a NLP-focused customer service representative called Contact Center AI. The AI takes incoming calls and uses NLP to offer solutions to customer problems. If the problem can’t be handled by Contact Center, the system will hand the call off to a human representative.

Also, Google isn’t the only company focusing on NLP. Recent reports show that companies like 3M, IBM, and Microsoft have all invested heavily in the NLP market, which is expected to grow by 16 billion by 2023.

If you’re still wondering whether or not NLP might not just be a passing fad, check to make sure a language-proficient robot didn’t write this article (it didn’t). However, ClickZ has used AI to write an article in the past, and if you’ve checked the sports report any time in the last couple of years, there’s a good chance you were reading a story written by AI.

The truth is, in the digital age where information is accessible in seconds without even touching our devices, customers expect answers quickly, twenty-four hours a day. Offering them robots that can communicate like humans is giving them the best of both worlds: personalized service without human limitations.

The post Ready to talk to robots? Natural language processing is changing search and customer service appeared first on ClickZ.

Original source: https://www.clickz.com/ready-to-talk-to-robots-natural-language-processing-is-changing-search-and-customer-service/216178/

Leave a Reply

Your email address will not be published.

List Building