It’s no secret to publishers (or readers) that placing ads around stuff people want to read is a deeply imperfect, still-evolving science. In the heyday of printed magazines, advertisers could make only broad assumptions about audiences. Who doesn’t recall flipping past pages of irrelevant ads and choking on cologne samples along the way?
Then came the Web, with its data-powered, keyword-targeted, user-tracking breakthroughs in efficiency and relevancy. But even today, a given Web advertisement commands only a fraction of the dollars that print ads rake in. The worst of those Web ads slide across the page as you’re trying to read, or start blaring audio from one of your browser tabs. And even the less obnoxious ones are slowing down our browsing experience. No wonder ad blockers are so popular.
But what if readers could have relevant, useful ads in a beautifully designed, nonintrusive format? The elegance of magazine ads with the intelligence of Internet ads: that’s what Flipboard is trying to do.
Today, Flipboard–the app that pretty much invented the modern, tablet-based magazine five years ago–is announcing a new ad-targeting system that promises more relevant advertisements to its millions of readers. Using what the company calls “interest graph targeting,” the new feature uses billions of data points about reader behavior to more smartly decide which ads to show to whom.
“For a long time, we’ve been really good at emulating the beauty of print ads,” says Dave Huynh, Flipboard’s head of ad product. “What this allows us to do is recreate that experience where the content and the ads are very, very closely linked. We’re replicating the value of the print model on a massive scale, but on mobile.”
About a year ago, Flipboard relaunched its social magazine app to focus more heavily on topics. That is, instead of just signing up, plugging in your social credentials, and picking your favorite publications, the app started letting users personalize their experience even further by selecting from more than 30,000 areas of interest. This way, even if none of your preselected channels on Flipboard tapped into your passion for typography or growing interest in international relations, you could still clue the app into these things and help it better understand what you’re looking for.
The data generated by this new topic-based system–combined with what Huynh says are “billions and billions of user signals”–has helped Flipboard build out this more detailed interest graph. This has helped Flipboard improve its content curation and discovery for the last year, but now the data is ready to help power ads as well.
And it’s not just about showing GoPro ads to photography or sports enthusiasts. By mining this massive trove of data about reader behavior, Flipboard is able to unlock more nuanced, sometimes counterintuitive, details about what readers are interested in. For instance, Huynh tells me, Flipboard’s data revealed a correlation between beer nerds and coffee enthusiasts–decidedly different beverages, but beloved with the same geeky enthusiasm. This is an unexpected, somewhat interesting connection, but more importantly, could it make an ad campaign for an espresso machine much more cost-effective. Don Draper would have killed for this type of intelligence.
Flipboard’s ads sometimes see click-through rates of 9%, according to Huynh. That may not sound like much, but it’s well above the average seen by Web display ads, which are lucky to go beyond 1%. Presumably, more smartly targeting ads by people’s interests should help boost the performance of some campaigns.
Flipboard’s interest-graph targeting doesn’t launch with any shiny, user-facing toolset. Rather, the team is quietly flipping the switch and letting its data flow more freely behind the scenes.
Five years after the launch of the iPad, tablets are not exactly the savior of publishers that many predicted them to be. But if mobile devices have any hope of helping the content business forge a more lucrative future, it will be features like this that help them get there.