Humans like to think we’re unpredictable. But in general, we fall into daily habits around our trip to work, the coffee shop we visit, the laundry detergent brand we buy.
Amazon knows this. That’s why it’s recommending products all the time — things you’re actually considering or hadn’t thought about but realize, yes, you want that. If you’ve recently talked about a product with a friend and then see it pop up later as a recommended item while shopping at Amazon, you might feel creeped out. But we humans also love convenience, and this predictability will save Amazon billions of dollars. So how does it work?
Data is sometimes compared to oil in its value. Whether you agree with that or not, information is making people money. Or rather, they make money from the ability to analyze mounds of data and view trends. Amazon has been gathering data about its customers from the start. This data includes:
- User profile – The one you filled out! This also lists your shipping address (location) and preferred payments, which includes credit cards.
- Computer – Amazon makes a note of your browser, IP address, time zone, and operating system.
- Purchases – Of course, they’ll note what you bought.
- Browsing history – Amazon keeps track of what you’re looking at, how long you spend on that page, whether you read reviews, click on things, etc.
- Searches – For what items did you search? What words did you use? Did you buy that item? Which and how many things did you click on before buying or not buying?
- Items customers recommend or review – Your reviews and the ones you read. If you love a particular item, that will affect future product recommendations, just as your negative review of something will.
- Abandoned shopping carts – Did you add items to your cart and then leave for awhile? Did you ever buy those or did you delete them later?
- Streaming – Now that you’ve got Amazon Prime, you’re also watching video or listening to music. What genres interest you?
- Kindle/Alexa/Fire Stick – Factor in the other Amazon products or services you use. Amazon is listening to you and recording what you read and watch. What questions have you asked Alexa lately? That may later tie into a product recommendation.
- Wish Lists – The things you add to your wish list are, presumably, things you want. Amazon also has a patent on technology that will track information about the people for whom you buy gifts. Now, it knows not only what you’ve purchased, but the things your mom gave you for Christmas last year.
Amazon combines all of this with other detailed information it may have gathered from other sources. Behold, a picture of your life, whether you are a 22-year-old recent college graduate in Raleigh who loves to watch period dramas and needs some new clothes or a 54-year-old ranch owner in Montana who doesn’t use all of his Prime benefits and may order a book. Amazon knows your practices and is using them for customer service. It also compares your habits to those of people who are similar, and often that’s where the predictions feel eerie. But although we all like to believe we’re snowflakes, another 36-year-old female homeowner in California who has bought gardening supplies wanted a hammock, so why wouldn’t you?
At the heart of this is Artificial Intelligence (AI), which is doing all the calculating behind the scenes. As AI via machine learning studies human behavior, it can pick out trends, especially with as large a data set as Amazon offers. It’s simple for the machines to see that other people in Chicago are buying hats in November when the weather gets colder and then provides you some options based on what people your age and gender prefer and what colors you have chosen in the past.
Amazon’s next steps are to anticipate purchases and begin shipping them before people buy, to reduce the time it takes from your purchase to delivery.
Many people are not sure they like Amazon, Google, and other companies knowing so much about us. If you’re bothered by it, you can go through what Amazon knows about you and clean out some of it. This is also affecting small and medium-sized retailers who do not have the power of Amazon to predict what customers will want next season. As Forbes outlined, merchants used to tell the customer what they wanted.
“They look at historical sales data from last season, assess the competition, and then use their own intuition to project what the consumer will want.” Now it’s the other way around, and retailers who bet wrong lose. “As of August 31, 2017, according to Retail Dive, 16 major retailers have filed for bankruptcy in the U.S., double the pace of 2016,” Forbes noted.
That’s why more retailers are focused now on the consumer and turning to AI and other tools to figure out what they want — even if the consumer doesn’t know.
If you’re curious about how AI can help your business, talk to us about solutions.