AI innovation finds a home on mobile devices

Mobile apps that anticipate

Amazon, and later Netflix, popularized recommendation engines that offer consumer suggestions as to things people might like to purchase based on recent purchases; for example, "People who bought a Schwinn bicycle also bought a Kryptonite lock." Online sellers are using the same idea to suggest what mobile shoppers might like to buy by using A.I. to analyze buying trends.

"I think where A.I. in mobile is most fascinating is its use as what I call an anticipation engine," said Tim Bajarin, who heads the market research firm Creative Strategies. "AI and machine learning are helping companies learn more about me, or individual consumers, to deliver personalized services that anticipate what I want."

It's not just e-commerce. "We're starting to get a taste of where this is headed when, for example, I put in my calendar I'm going to meet someone at a certain restaurant at noon and an hour before I leave a map to the restaurant pops up on my mobile device," said Bajarin. "Or another example, I can tell a bot on my phone before I head to the office that I need to research a certain topic. By the time I get there, I've been sent 20 links on the topic and three summaries."

Winning over customers

During a panel discussion on "delivering mobile customer engagement," executives from online vendors and payments company Square, said A.I. is helping them attract and retain customers.

Saumil Mehta, customers lead at Square, said the more than two million sellers using the company's point-of-sale device and services on mobile phones and tablets get more buyers spending more money thanks in part to the data it collects. He argued that for company's it's important to collect customer data from the start -- even if you won't be using it right away.

"The first thing we decided at Square was that every time a credit card is swiped or tapped we update the customer profile behind the scenes," said Mehta. Later, as the database of customer profiles grows, Square can offer users things like automated loyalty programs. They include, for example, customers who spend US$500 and get US$50 off and "Win Back" campaigns that automatically contact customers who haven't bought anything in 90 days with special offers.

Nikihil Raghavan, vice president of product for search and machine learning at Etsy, said the online seller of jewelry and other hand-crafted items uses AI to help manage billions of search queries and offer personalized services.

"We're using deep learning to figure out what buyers tend to find interesting, and part of that is using recommendation technology to figure out what are the right items to send you," said Raghavan, who was co-founder and CEO of Blackbird, a deep learning startup that Etsy acquired in 2016.

As for mobile, Raghavan said it can be a challenge for online sellers given the more limited screen real estate than desktop and notebook computers offer. "You have to intelligently structure the user experience on mobile, and get creative on the UX side," he said.

Computerworld (US)