IBM Watson might be best known for outsmarting "Jeopardy!" champion Ken Jennings to win $1 million on the game show, but this year the supercomputer system has a new goal: to be the best Santa's helper ever.
IBM (IBM) has created a new application called IBM Watson Trend, which sorts through millions of online conversations and predicts which gifts are trending up (and which are out) before they actually reach peak demand. The issue isn't a small one, as anyone who can remember the Cabbage Patch Kids shortage in the early 1980s can attest.
IBM Watson Trend, which is available for free to consumers through IBM Watson's website and through a downloadable app in the iTunes store, sets itself apart from other lists of hot holiday products because it's mining data on the sentiment consumers are expressing about products they've bought or want to buy. That can help predict whether a gift will drop off consumers' radar or if it could become the next Cabbage Patch Kid -- desired by everyone, but nowhere to be found.
"We realized we could take the power of Watson and understand whether or not people are buying, but also understand what's behind the trends and why people are buying," said Jay Henderson, director of IBM Marketing Cloud.
One trend Watson has identified isn't too far off what noncomputer-based trend-spotters (also known as humans) have identified: Star Wars toys are hot this year.
What's the difference between what a human can spot versus Watson? The IBM app is geared toward providing a forecast for demand, as well as figuring who is buying the product and why.
In the case of Star Wars-themed Lego sets, Watson is predicting that three types of sets are likely to sell out: the new Millennium Falcon, Lego City, and Friends. Therefore, the artificial intelligence is recommending that shoppers take advantage of early Black Friday sales and score their deals before running into any out-of-stock notices.
That's something Henderson said he took to heart after noticing that IBM Watson was predicting an increase in demand for Minecraft-themed Lego sets.
"My son is a huge Minecraft fan and a huge Lego fan," he said. "On his Christmas list were some Minecraft Lego sets, and when I saw this trend I actually bought them online line that day."
Other hot trends Watson is picking up include Samsung TVs, given high ratings from current owners (although men are more interested in them than women), and the Apple Watch, which is generating buzz, thanks to release 2.0.
Another idea Watson has generated: Look for a stand-alone digital camera for your friend or family member who loves taking snaps on their smartphone but wants an equipment upgrade. Nikon digital single-lens reflex (DSLR) cameras are coming out as the top choice for this group, followed by Sony's 42-megapixel Alpha 7RII mirrorless camera.
Barefoot running shoes are also gaining speed this holiday season, according to Watson. It's finding that women are starting to buy them as fashion statements, and not necessarily for their performance edge, Henderson said.
But with Watson predicting which trends are hot or not based on millions of social media mentions, blogs, forums and ratings, won't that lead to self-reinforcing trends, similar to how the rich get richer? Henderson said he doesn't believe that will be the case.
"We're hoping to achieve the opposite, by connecting consumers to information and a lot of different trends," he said. "Rather than everyone buying the Cabbage Patch doll, you'll have people buying the cameras or the sneakers for the fashionistas. It'll enable each individual to find the products that fit them the best, rather than one big product for the holiday season."