• by Steve Smith , Staff Writer @popeyesm, December 9, 2013

For as long as I have been covering digital media -- especially behavioral targeting -- I have also been watching the promise of personalized content and services go largely unfulfilled. Besides MyYahoo, a very successful personally targeted home page for the portal, and e-commerce experiences notably from eBay and Amazon, most providers have fallen flat in their attempts to target digital content to specific tastes.

Personalization was the rage into the middles of the 2000s, with personalized search engines of all kinds, personal portals that tried to aggregate news around users' browsing habits, and major media sites that surfaced personal preferences in a column or two of the main page. In the end, though, many publishers lamented the fact that only a small percentage of users opted into customizing their own experience. That's when the content recommendation engines kicked in and started targeting other content to the user, in a personalization-lite model.

Arguably, it is with social news apps like Flipboard, Zite and Pulse that we started seeing a more sophisticated approach to combining active and passive behaviors to curate content. Once users declare interest in a vertical, the behavioral or social media engines take over to prioritize specific kinds of content most likely to appeal to them. This model works for aggregators, but not as well for major media. For the big content brands, personalization always presented a conflict with traditional editorial functions of curating news for readers. An over-personalized experience could depress the content the publisher most wants to circulate,  and cut the serendipitous experience of discovery.

One way to overcome these problems is to personalize the user experience on the messaging end rather than at the site. For instance, a personalized news app I don’t even use anymore, Trapit, has my preferences and my email. I get regular notices from them with top articles surfaced from my interest areas. Real Estate entities like Realtor.com are converting my recently saved home perusals into email notices of nearby similar properties.

In many respects the app world is leading us into a next generation of content personalization through messaging. ESPN recently revised its SportCenter app to allow for much more customization by favorite teams. But the brand has also been a pioneer in highly targeted alerts for new scores and stories related to favorite teams. Similarly, People magazine’s CelebWatch app lets stargazers opt into alerts for specific celebrities.

Messaging is the killer app for apps in much the same way that email has been for Web sites. But the impact can be even greater. App message platform Urban Airship claims that one of its partners, the London Olympics, got 60% click-through rates on 10 million location-targeted messages, a 10X higher rate than emails. It almost goes without saying that users who opt in to push messaging are retained by an app at a much higher rate. Across Urban Airship’s customer base, only 29% of people who did not opt into messaging even reopened the app within the first month of downloading. But 55% of those who subscribed to alerts did reopen the app again.

Granted, these are figures coming from an app alert platform, and the user who opts into messaging is already the most likely to revisit the app. Still, this only underscores the importance of publishers emphasizing the messaging part of their value proposition. Getting a user into the messaging loop improves retention long term and certainly improves engagement overall. Across verticals, alert recipients opened an app on average 35% to 41% more often. And for some apps, the messaging becomes the personal front end to users, responsible for up to 83% of opens.

At Urban Airship, in the retail vertical 46% of users opted in for messaging, while in the media category 50% did so. This last metric is especially revealing because it suggests how much alerts are content. Rather than clicking through to open the app and read the story, it appears many people rely on the alert to give them all they really want to know. But users seem to like it this way. Half of those who opt in to alerts in media apps are still using the app six months later, compared to 31% who opted out. Publishers may not be able to monetize messaging on the front end, but it is a part of an overall content and communications strategy that keeps the user enveloped in the brand.

By the way, the push messaging model is not just for devices -- or at least it won’t be for long. Apple has already extended this capability to its Mac Safari browser, and there are a number of APIs out there that bring different flavors of Web notices to the desktop. One hopes there will be some standardization here to allow users to opt into site notifications across OSes and browsers. But the app is the leading edge for now, and suggests how push personalization could help turn old models of content into conversations.