As some industry insiders speculated, IBM is starting to merge its Weather Co. asset with its cognitive supercomputer Watson, a move that will benefit its marketing cloud and commerce businesses.
“Everyone wondered why an IT company like IBM would make such an acquisition, a company that produces forecasts for 2.2 billion locations every 15 minutes,” said Harriet Green, VP and GM of IoT and commerce for IBM, at IBM Amplify in Tampa, Fla., on Tuesday.
“We’re the second largest location services company now in the world. Weather affects people’s mood and how much they’re willing to pay.”
While IBM is still in the integration phase, its Marketing Cloud already enabled weather-based triggers.
But eventually it will merge weather-related triggers with a marketer’s broader digital assets and campaign creative, and ultimately predict the impact of a “weather event” on future sales.
As an example, a bike retailer using Watson and Weather could define attributes for a “commuter bike event” (translation: a consumer shopping for a bike to ride to the office) that will automatically populate text and images for ads related to the current weather.
Watson might identify “4,863 locations with below-average temperature or greater than 50% chance of precipitation” so marketers can decide if if they should update the campaign by adding content related to bad weather.
Watson might then consider past CRM or sales data, social media and third-party data sets to suggest other “relevant attributes” for targeting in that respective (or future) campaign.
But that’s just the targeting part. IBM claims cognitive learning [e.g., Watson] will help move the needle on market forecasts for future demand for a product, which extends well beyond marketing.
“I think we’re just starting to make the connections between IBM’s Universal Behavior Exchange [in marketing] with the commerce side of the business,” said Chris Victory, VP of strategic partnerships for MediaMath, a UBX partner.
“There are a lot of interesting applications for Watson to make recommendations off of disparate, structured data sets, such as product feed or clickstream data” and layering in Weather’s location-based data on top of it.
Forecast On The IBM Cloud
IBM says its analytical approach differentiates it from its marketing cloud competitors like Oracle and Adobe.
“There are companies that are about data itself or the creative process, but IBM is and always has been about helping clients gain insights from analytics, which is why we’re laser focused on cognitive,” said Steve Mello, VP and business leader for ecommerce and merchandising for IBM Commerce.
Although cognitive marketing may be a couple of years out until it’s adopted at scale, IBM was laying the groundwork for what it perceives as a major market opportunity.
“We bought Weather because it was an amazing platform that can merge together millions of transactions and events, not just for the fact that it was the no. 2 app after Google for global positioning,” Maria Winans, CMO of IBM Commerce, social and mobile, told AdExchanger.
Winans said the lifespan of a new product or capability typically takes about 18 months to get from ideation to a ship-to-market stage.
“You’ll start to see us embed cognitive capabilities, such as Real Time Personalization,” she said, “where [weather data] may be [one data feed] and where we come up with visual recommendations for ads.”
Buyers seem excited about the IBM’s “cognitive” pitch, particularly the intersection of location-based data and advanced analytics.
Tessa Horehled, VP and group director, digital and social strategy, at DigitasLBi, called the current predictive modeling solutions that are in-market today “heavy” and “conceptual,” but thinks IBM’s ability to combine location data and cognitive learning will be much more tangible for marketing use cases.
“A huge theme at IBM Amplify this year is the customer journey – gaining full visibility into the omnichannel journey,” she added. “For example, a brand can leverage IBM’s Universal Behavior Exchange within the IBM Marketing Cloud to capture behavioral data from sources such as e-commerce, marketing automation, web pixels, and mobile SDKs, move the key behaviors into the social marketing solution, Sprinklr, and apply this data to their new paid social ad campaign.”
