Dive Brief:

• IBM unveiled a suite of new artificial intelligence (AI) marketing solutions at Advertising Week 2018, including Watson Ads Omni, an iteration of Watson Ads that enables marketers to deploy AI-powered interactive placements on any digital site. Lego Systems will be the first brand to use Watson Ads Omni on Black Friday, according to news shared with Marketing Dive. The ad format was trained on 35 different Lego products and allows the toy brand to create tailored responses to consumer interests and needs.
• Another new tool, the IBM Media Optimizer, leverages MediaMath's cloud-based demand-side platform (DSP) and data management platform (DMP) with a goal of helping to optimize programmatic media spend. By leveraging Watson AI, the media optimizer also aims to help marketers centralize their paid and owned customer behavior data within an integrated platform that allows for efficient campaign measurement and adjustment. The tool includes something IBM calls Intelligent Bidder, which uses AI to improve programmatic paid media spend and impact by reducing cost per acquisition. Intelligent Bidder does this by analyzing the time, channel and bid value for each ad placement for each audience based on conversion probability in real time. 
• IBM at Advertising Week also introduced Predictive Audiences, a new targeting capability leveraging deep learning with neural networks and LiveRamp's IdentityLink. The tool intends to help marketers develop models of their target consumer that will offer more precise targeting and contextual relevance. The solution layers in continuous data signals, like location, weather and first-party CRM data, with third-party data from the LiveRamp Data Store in order to identify consumers that would be the most or least likely to take a specific action. 

Dive Insight:

IBM's new Watson-powered offerings focused on personalization, campaign targeting and optimization arrive ahead of the critical holiday season, where sales are increasingly focused on channels like e-commerce. With tools like Watson Ads Omni, IBM is looking to help brands cater to more specific audiences — in Lego's case, moms looking to buy a toy set for their children — by sourcing more information directly from the consumer and analyzing that data through AI.

"Moms will be able to say, 'here [are] my kid's passion points' — they love dinosaurs, they love pirates ... and what Lego's building for them, through the use of AI and a bunch of different, layered targeting pieces, [is] a recommendation engine," Carrie Seifer, Watson Advertising's chief revenue officer, told Marketing Dive at Advertising Week. 

IBM also announced that Behr Paint is the first retail brand to leverage its interactive Watson Ads, which launched in 2016 at Advertising Week. Other categories, including automotive, healthcare and CPG, have already seen success with the format. 

The lawn care company TruGreen, for example, began using Watson Ads in April to provide personalized planning recommendations, answer questions about lawn maintenance and connect consumers to experts. Best Western also partnered with IBM on an AI-powered campaign to offer vacation planning and travel tips, accommodation recommendations and offers. Both brands' ads ran on The Weather Channel app and weather.com, which IBM owns. 

For Behr, the company is combining Watson's Natural Language Processing and Tone Analyzer capabilities to offer personalized paint color recommendations based on the mood or feeling a consumer wants for their space.

"You layer in different data sources like Behr's product line, social sentiment, information that the consumer's giving you as they interact with the ad ... that [data] starts to get a little bit thick," Seifer said. "AI will come in and look at all the data in real-time and say, 'this is the paint, this is what you should decorate with.'"

Peter Adams contributed reporting.