Short answers, pulled from the story.
Targeted advertising, also called data-driven marketing, directs ads at audiences with specific traits including demographic, psychographic, or behavioral characteristics. Advertisers use tools such as HTTP cookies, web beacons, and data mining to build profiles of users based on browsing history, purchase behavior, location, and social media activity, then serve ads matched to those profiles.
A 2009 Network Advertising Initiative study found that targeted ads generated an average of 2.7 times as much revenue per ad as non-targeted run-of-network advertising and converted clicks into purchases at twice the rate. Research by Farahat and Bailey on Yahoo data found the marginal cost of a brand-related click dropped from 72 cents without targeting to 16 cents with it.
Targeted advertising requires the aggregation of large amounts of personal data, including sensitive information such as health issues, sexual preferences, and precise location, which is then shared among hundreds of parties during real-time bidding. A 2012 Pew Internet and American Life Project survey found that 68% of Americans said they were not okay with targeted advertising because they did not want their online behavior tracked and analyzed.
Research cited in studies of targeted advertising found that gender targeting is accurate only 42% of the time, which is worse than a random guess. When targeting by both gender and age simultaneously, the accuracy falls further to 24%.
In December 2023, 404 Media reported that Cox Media Group had been marketing a service called Active Listening to advertising professionals, claiming it could access microphones on smartphones, smart TVs, and other devices to target ads. Google removed Cox Media Group from its Partners Program after a review, and Cox Media Group removed the promotional material from its website and denied listening to any conversations.
The main types are demographic targeting, which segments audiences by age, gender, income, and similar traits; behavioral targeting, which tracks browsing and purchase history; contextual targeting, which places ads next to relevant content; psychographic targeting, which uses personality, values, and lifestyle; location-based targeting using IP addresses and mobile app data; and retargeting, which serves ads for products a user has already viewed.