A new app leverages the power of IBM’s Watson, gleaning information from more than 100,000 sources to create an accurate predication of holiday sales. Social media buzz, comments, and reviews are all fair game when it comes to prophesying what shoppers will actually buy. (IBM Watson’s biggest claim to popular fame thus far is its win on the trivia show Jeopardy!, where it beat two human competitors for a $1 million win.)
Available for iOS and desktop, the IBM Watson Trend app’s initial purpose is consumer-facing: to help shoppers discover this year’s hot trending products, find out where they are available, and plan their holiday shopping around that info. But eventually, the plan is for a version of the app to be used by enterprise as well.
This could be an interesting year for the pilot of this app, as various forecasts are predicting that the holiday sales will be slightly flat compared to last year: The National Retail Federation predicts that holiday sales through December will increase 3.7 percent, less than 2014’s 4.1 percent.
On the other hand, some argue that Black Friday shopping predictions aren’t very accurate and should be ignored. Everyone wants to end the year on a positive note and with good sales figures, so the temptation is strong to put a definitive market on something that’s not definitive (at least until after the fact, and by then it’s January). Initial information is based on things like informal shopper polling at malls, and one weekend of sales doesn’t lend itself to overall predictions of shopper moods and what’s likely to happen over the coming weeks.
So predictive analysis, which has been used to uncover shopping trends and consumer behavior for a few years now, could become even more important as new and interesting uses of this technology are developed.
One example comes from the Pabst Brewing Company, which is using predictive analytics to change timelines, accounting, and inventory. The company has even gotten rid of annual budgets and switched to a quarterly, rolling 18-month forecast.
Grocery store chain Kroger is also using predictive analysis to predict when shoppers visit its stores. The company can use the information to alleviate lines and better schedule cashiers, leading to a more efficient check-out experience (and likely happier customers).
The benefit of using social media for predictive analysis is immediacy and volume. But the difficulty remains of applying that data so that it’s useful.
For more insight on how analytics can boost business, see Analytics: The Engine Room Of More Successful Sales And Marketing.