We all love a good internet browse. With tablet technology now ubiquitous, it’s becoming the norm to wile away the time bouncing around the web. But the amount of time actually spent enjoying something new and interesting is often eclipsed by the time it takes to find it. The sheer volume of content produced is growing at an enormous rate. According to recent stats, in December 2010 the web contained 225 million websites, 21.4 million of which were new in 2010. There are an estimated 152 million blogs online and 25 billion tweets were sent in 2010, many of them in turn linking to more content. According to recent research by the University of Southern California, each person is bombarded with the equivalent of 174 newspaper’s worth of data a day. With all that content being produced, all instantly available online, the challenge for most of us has shifted from content accessibility to content filtering and curation.
Companies who publish content for a living are now adjusting to this new challenge. For the last few years, there has been an emphasis on creating more content more quickly, feeding an underlying assumption that if you build it they will come. But with so much competition, and given the resources it takes to continually create massive amounts of high quality content, many publishers now realise that simply creating it is not enough. They are investing in ways to break through the noise. This process involves honing techniques both to attract new audiences and, as important, to keep audiences on site, consuming more and more content once they’ve arrived. After all, with so many choices, patience runs thin and people will quickly leave a site if they don’t discover something of continued interest instantly, and within the flow of their experience.
One tactic publishers have used to help people find more content, and stay on site longer, is to link to additional articles they think the person may want to read. However, early approaches to identifying what someone would want to read have historically been too simple and generic to make a meaningful impact. For instance, publishers often list links to their most popular articles, but this often just shows people what they have already read (that’s how they became popular in first place!). Or publishers will present their most recently published content. But with the exception of a breaking news site, to-the-minute timeliness does not usually trump personal interest. Another more sophisticated tactic, fuelled by the advancements of search and contextual technologies, has been to automatically provide people with links to ‘related content’. But while a definite improvement, even this approach often fails to engage readers — most people browsing content are not actively searching for a specific topic and don’t care to read pages and pages about the same subject.
So what can publishers do to attract and re-engage users, allowing them to discover content that’s truly interesting to them? At Outbrain, we’ve been fortunate to be able to test a variety of methods across thousands of publisher sites, working constantly to develop new ways of identifying what content is interesting to different audiences. And we have learned a few key points along the way. First, it’s important to personalize your content offerings as much as possible. For instance, the simple step of removing links to articles a person has already read helps to keep them looking at your recommendations, knowing they will always find something fresh and new there. Second, leveraging information about what groups of people find interesting can help you pinpoint what an individual will find interesting: relationships between people and their reading habits are more useful data points than topical relationships between articles when trying to surface content interesting to each user. And third, filtering for content quality is an important tool, and it is a very different task than filtering for content relevancy. Assessing quality based on user engagement data, and then giving more emphasis to those pieces of content that have proven interesting to others, helps you whittle down an overwhelming amount of choices to a smaller subset more likely to be valued by people who encounter them.
There’s no silver bullet to solve the challenges of information overload and content discovery. People continue to advance their own personal filtration methods: moving from opt-in email newsletters to RSS, from accessing socially driven sites like Digg to now following helpful ‘editors’ on Twitter or reading stories posted by friends on Facebook. But the publishing community, too, has an equally important role to play in making it easier for people to find those perfect stories when they visit their site. I, for one, am excited to see the advancements being made in this area and look forward to spending less time searching, and more time reading.