• by Laurie Sullivan , November 18, 2015

The group at IBM Watson released an app Wednesday that enables consumers to keep up with trends before buying gifts during the holidays, but the opportunity for marketers is in the data to double-check holiday promotions, determine features on the home page, and forecast demand for products.

In typical IBM fashion, the app was born from an internal brainstorming and innovation session brought to life with Justin Norwood, offering manager for IBM Watson Trend, leading the group. "We have aspirations for marketers to use the app, but also consumer products companies that need foresight on trends to innovate," he says, calling the first version a gift-giving guide. "Perhaps in six to 24 months, it can really start to inform other parts of the company like merchandising and supply chain."

IBM's cognitive computing app predicts the hot electronics, toys or other items that will likely sell out by a specific date. It provides trend scores and details behind the trend. In fact, users can view the top 100 trending products across three categories -- consumer electronics, toys and health and fitness.

The data is pulled from "tens of millions of conversations around purchase intent, purchase experience, and post-purchase emotion," from more than 10,000 Web sites like social media, top 30 blogs, ratings and reviews, and ecommerce sites.

The data also comes from search results, but Norwood declined to elaborate on which engines or sites. The Watson technology not only understand keywords -- similar to search data -- but meaning, context and tone, he says. "Sometimes search data can be a bit misleading, because if I'm searching for the iPhone 6, the search engine doesn't know if I'm planning to buy an iPhone or I'm angry because of a flaw because I'm looking for warranty information," he says, acknowledging that Watson suggests why a product is trending.