It’s a complicated thing to explain but TechCrunch does a decent job describing ad targeting company RadiumOne’s new feature, enabling RTB advertisers to bid on users who have posted or viewed content with hashtags.
Whenever a user publishes a hashtag, or even views a piece of content labeled with a hashtag, the site can drop a cookie about that hashtag onto the user’s browser. Then when the user surfs around the web, they’ll end up as one of the 25 billion impressions RadiumOne sees each day across sites that host its retargeted ads. Those sites detect the hashtag cookies on the users’ browser that indicate they’re interested in #fitness or #justinbieber. The site pings RadiumOne, whose real-time bidding platform lets advertisers like Under Armor or Ticketmaster compete to reach those users. The highest bidder gets their hashtag retargeted ad shown, and RadiumOne splits the revenue with the site that dropped the cookie in the first place and the one hosting the ad.
RadiumOne is doing this now on its own site, Via.me, but it could expand the capability to other sites.
Here’s a link to a video featuring a Performics Asia-Pacific exec, Gareth Mulryan, talking about how Performics supports real-time performance analysis. Key tidbit:
“Achieving real-time performance hinges on your ability to collect and quickly derive insights from large volumes of participant and cross-channel data.”
Not long ago, I attended the Seattle Interactive Conference, a two-day gathering of the best of Seattle’s tech minds. One of my favorite sessions featured Quantcast CEO Konrad Feldman, talking about the intersection of big data and advertising.
The amount of data that Quantcast is dealing with is incredible; Feldman said the company works with 40+ terabytes of raw data each day. Marketers need to shift their thinking, he said, away from the old model of media buying (based on impressions bought in aggregate) and toward the model supported by real-time bidding (impressions bought and priced individually).
Here’s a link to his deck on Slideshare.