AI Case Study
Rocket Fuel uses NLP to protect brand reputation when determining online ad placement
Rocket Fuel in collaboration with Discovery News Service, both based on IBM's Watson, has automated ad placing. Rocket Fuel's predictive marketing platform leverages Watson to score the sentiment of news headline and evaluate it in relation to a brand. Discovery News AI's outputs serve as input signals to the Rocket Fuel AI, thus achieving AI to AI communication, to ensure that ads are displayed on pages with positive stories about the brand or persons associated with it. As a result, brand marketers are able to protect their brand identity and reputation.
Industry
Consumer Goods And Services
Media And Publishing
Project Overview
"By integrating AI-to-AI communication between Rocket Fuel and IBM Watson, our Predictive Marketing Platform scores the sentiment of breaking news headlines and responds accordingly, reducing the risk that brands will appear on the same page as a negative news article. Using Watson, our brand marketers can now efficiently assess the positive or negative sentiment impact for a brand across news sites, and control — at the page level — where their ads are shown.
A marketer inputs brand names, product names and celebrity names that the brand has identified as important to their campaign. Every five minutes, Rocket Fuel queries the Discovery News service, instructing it to find articles with these names.
Meanwhile, Discovery News consumes more than 200,000 news articles each day from multiple sources and, using Watson natural language understanding techniques, analyzes the sentiment of those stories (i.e., on a scale of +100% positive to -100% negative).
Upon receiving the query request from Rocket Fuel, Discovery News returns the latest articles that contain the brand, product and celebrity names along with the publisher source and the sentiment score. The results are then displayed on a user-friendly dashboard within the Rocket Fuel platform, where brand marketers can view the results and choose keywords accordingly, giving both marketers and clients peace of mind.
Consider a brand that has a well-known celebrity endorser – for example, Super Bowl winning quarterback Tom Brady. Showing ads for its products alongside articles that mention Tom Brady in a positive light can amplify the connection and provide additional product exposure. Watson Discovery News analysis of one article, “What young people can learn from Tom Brady’s Super Bowl Win,” results in a high sentiment score and activates a positive keyword phrase “Super Bowl Win” in the Rocket Fuel platform. The Rocket Fuel AI can now use this positive keyword in its scoring algorithms.
Conversely, another story in which Tom Brady’s wife, Gisele Bundchen, talks about his history of concussions may not be a suitable place to show the brand’s ad due to the ongoing controversies around football player safety. This article would receive a negative sentiment score and trigger negative keywords (“Concussion, “Football Player Safety”) in the Rocket Fuel Platform. The brand manager would be able to specify how the Rocket Fuel AI would treat bid opportunities with these negative keywords in the web page URL, including ignoring the keyword, blacklisting the URL, or pausing the entire campaign.
The Discovery News AI is embedded within our platform such that the Discovery News outputs now serve as input signals to the Rocket Fuel AI. We connected two intelligent AI services directly through API-based service calls."
Function
Marketing
Digital Marketing
Background
"Rocket Fuel is a predictive marketing software company that uses artificial intelligence to empower agencies and marketers to anticipate people’s need for products and services. Rocket Fuel was founded in March 2008 with a vision of transforming the digital advertising industry through the use of big data and artificial intelligence.
Each day, Rocket Fuel Inc. evaluates over 200 billion opportunities to determine when and where to place an ad on behalf of its brand clients.
In this era of concurrent news cycles, negative news can erupt out of nowhere. It’s getting harder and harder for brands to ensure brand safety on news sites. You could blacklist the publisher, but that could lead to negative repercussions. For example, blacklisting the Washington Post or USA Today to prevent your ads from appearing on a page with a negative news story could also result in blocking your ads from being seen on a page with a positive news story about your brand."
Reported Results
Results undisclosed
Benefits
Technology
Data
Online digital information including news articles