So, you had a baby a couple of years ago, and you go to a store’s app to search for a toy for your now-toddler. And whaddayaknow, there's a sale on the perfect treat.
That could become more common as artificial intelligence continues to creep into our mobile shopping experiences.
One Pittsburgh company, CognistX, is at the forefront of that movement. Its mobile app enhancements let retailers use advanced information about a person’s lifestyle and spending habits to target specific content toward shoppers.
“You are going to the mall and you maybe have not shopped in a while, so what is the right offer that we can present to you that’ll drive you to go to a particular retailer’s store and say, ‘This is exactly the dress I was looking for, or accessory I was looking for?’” co-founder and CEO Sanjay Chopra said. “Or how did they figure out that my car needed brakes or an oil change? And you send the right promotion at the right time and drive profits.”
Co-founder and Carnegie Mellon University computer science professor Eric Nyberg said these companies are already collecting some data, but CognistX’s artificial intelligence is able to sort through it for them.
“Retailers have the data, and they don’t know what to do with it,” he said.
Those retailers are able to infer loose associations between a customer’s lifestyle and spending habits, but artificial intelligence recognizes fundamental correspondences between shoppers' characteristics and where they spend their money, he said. It’s something CognistX’s team has been working to perfect over seven years.
“Machine learning technology, which is a big part of artificial intelligence, is so important in the retail context,” Nyberg said, “because I could sit and look at 100,000 transactions and not necessarily make much sense out of which customers prefer what products.”
CognixtX enables retailers to track past purchases, but also recognize trends that predict what someone might buy next. Maybe you’re eating healthier now than you did a few years ago, so you’d love a deal on produce or workout attire. Or you were buying diapers two years ago, so it offers products aimed at toddlers.
Retailers can also track social media, web searches, what time of day someone shops and his or her relative location. That may sound like an invasion of privacy to some, but Nyberg said it can help a customer feel valued.
“You can personalize the behavior of a cognitive system right down to the level of an individual so the individual feels like you’re really responsive to their context, their history, their interests,” he said. “And it’s very difficult to think about doing this without statistical machine learning.”
Chopra said he estimates a 370 percent return in investment for companies who use CognistX.
He said he's talked to representatives of clothing companies and sporting goods stores about how the app can increase profits. Monroe Muffler and Brakes, Mr. Tire and Giant Eagle supermarkets are already hooked, he said.
For some, Chopra said the biggest appeal comes from reducing redundancy in advertising.
“So Giant Eagle will send a flyer out, and all of us get the same offers,” Chopra said. “They don’t figure out if you’re a vegetarian or a vegan or a gluten allergy. They have no idea.”
CognistX knows. Nyberg said the information sharing is an opt-in process because a person has to download the app in order for the data, like browser history and GPS, to be collected from a phone.
“So I think there’s that inevitable concern about whether this is creating Big Brother,” Nyberg said. “Is it going to gather all this data about everybody, and then use it for some nefarious purpose? But obviously we’re only going to be gathering data that the user wants to share with us about their retail behavior.”
In this week's Tech Headlines:
- University of Pittsburgh medical researchers think they know why some women experience more severe symptoms of menopause than others. The study by Pitt's Graduate School of Public Health examined the physiological causes of hot flashes and night sweats. They determined women who are obese or suffered from depression were most likely to experience an early onset of symptoms while smokers, women who maintain an ideal weight and African-American women tend to experience these symptoms later in life. Chinese women were least likely to have severe symptoms.
- Those little emojis on your smart phone are about to undergo another update. Unicode Consortium, which controls emoji standards, has agreed to add 11 new emojos with the goal of highlighting women in diverse careers. So starting next year, look for a female doctor, scientist, farmer and welder within the emoji library.
