Have you ever been reading a serious article about troubles in the Middle East, but somehow find yourself bombarded by irrelevant recommendations for articles about Justin Bieber or how to “lose weight fast?”

Does this feel off-putting to you? It should.

For publishers to build a loyal and engaged readership, they need to understand their own content inside out to offer users valuable and relevant recommendations that motivate them to engage, read, and share content.

Publishers must scratch their heads when they see traffic pouring into their site, only to see a staggering 75% of users leave after only one page view. This bounce rate reflects the short attention span of digital audiences, particularly on mobile, but also highlights the publisher’s inability to grab its target audience when they are already on its site.

Many publishers have made a half-hearted effort to engage users via social media channels. This may drive traffic to individual articles, but it’s rarely loyalty building.

More likely, it will be another “one-and-done” page view where a user clicks the link, goes to the recommended page, reads some of the article, and returns to where they came from without exploring the publisher’s site.

This kind of “a la carte” viewing is driving page view numbers, but not at all building loyalty or readership.

Let’s say a user ends up on a publisher site after clicking on an acrobatic kitten post on Facebook while on his/her mobile. Predictive data should be able to ascertain that this user is not in the mood for article recommendations about the death toll from the Ebola crisis.

Data should inform content recommendations to serve up something suitably targeted to that user’s mood at the time s/he arrives on the publisher’s site.

Content recommendations need to cater to context. And a publisher should have a solid understanding of what content exists on its site(s) and what would be appropriate to recommend at a particular time and context.

The user who arrives after clicking on the acrobatic kitten might be more interested in news about George Clooney’s recent nuptials than the recent ISIS beheadings.

Context works both ways.

The user who arrives on a publisher’s site following a link about Syrian refugees is not likely to be interested in reading about Miley Cyrus’s latest publicity stunt. It’s all about gauging user mood and interest and acting on it: The right content at the right time.

Other publishers specifically enlist “click-bait” content, recommending gossipy or gimmicky articles from content networks that not only steal your user data, but also your dignity.

In many cases, this backfires in a big way, as the publisher devalues its own brand, especially if it is a premium publisher, such as a credible news site. Click-bait not only drives away many potential new readers, but often alienates existing loyal users who rely on quality content that support the publisher’s brand identity.

What may seem like a failure to capture audiences is actually a big opportunity for publishers.

Publishers realise that more sophisticated methods are needed to deliver meaningful content recommendations that address individual preferences and strike a chord at a particular moment with an understanding that this is the basis for building lasting user relationships.

The winning team at a Global Editors Network (GEN) “Hackdays” event, BBC, developed a tool to help publishers build loyalty with visitors who come to their sites on mobile from social media.

The BBC project used Cxense analytics and content solutions to help publishers cater news to readers’ habits. The solution led to more relevant content recommendations based on whether readers enter through the site’s home page or if they are referred by platforms such as Facebook.

The BBC project was award-worthy and timely because it is critical that publishers understand their audience and their own content and how they interact.

Growing the user base and keeping it loyal is a matter of being able to act on real-time user information across all devices to capture users while they are still on the publisher’s site.

While this has been easier said than done in the past, technology is now available to help publishers understand their audiences and content – and then leverage that information to make intelligent content recommendations that build their brand rather than break it down.