5 ways startups are narrowing the content discovery gap

Consuming content on the Internet has long felt like drinking from a firehose. Yet the problem is getting even worse. Driven by a focus on page views and the growth of viral content from social sharing sites like Buzzfeed and Gawker, the volume of unfiltered and even unwanted content has exploded. Users continue to struggle to find the signal in the ever-louder noise.

In addition to wasting users’ time, this content discovery gap poses a significant opportunity cost for those who trade in information currency: publishers, analysts, investors, entrepreneurs, executives, journalists and other professionals. Like the adage “half of my advertising is wasted – I just don’t know which half,” the vast majority of content misses its target and goes unread, creating a vastly underperforming asset for producers and publishers. Fortunately, startups and innovative publishers are developing and employing strategies that promise to narrow the Content Discovery Gap.

Curation: A partial solution

Publishers have always “curated” content for audiences, but as the volume and variety of content have expanded, presenting only relevant content to readers has gotten more difficult. A number of startups are betting they can improve on curation – both News.me and NewsWhip, for example, highlight stories by topic that are “spreading fastest” on Twitter. Inside is relying on brevity (approximately 300 characters per summary), clarity (conveying key facts), coverage (more than 1,000 stories a day), craft (individuals trained in the “art” of curation), and customization (readers can select from 175 topics, ranging from “Advertising” to “Yahoo”).

Despite numerous attempts over the years, curation platforms have not attracted a wide following, in part because topic categories (e.g., “advertising”) are too broad for users to effectively filter out unwanted content. In addition, even though the results are abbreviated, they are still a stream. The problem with any stream is that it’s difficult to match content to a reader’s interests at a particular point in time. While curation tools have value, their success depends on how well they learn and match users’ interests.

Chunking: Bits of content, details of engagement

Publishers and curators have always relied on headlines, captions, images and abstracts to convey details of a story to readers that will, ideally, entice them to read it. Mobile app Circa applies a variation of this approach to the content itself, splitting a story into its atomic elements (e.g., facts, quotes, images). With a Flipboard-inspired user interface, Circa allows the reader to swipe to view the next chunk of information in a story. Besides allowing readers to easily get the gist of a story, these smaller, more digestible bits allow Circa to track how much of a story someone has read, revealing the user’s level of interest in a story or topic.

Chunks fit well on smartphones and small-screen tablets, but because users must swipe or click to advance through a story, they are less efficient for users on a laptop or larger-screen tablet. From the publisher’s perspective, chunking provides more granular measures of users’ interests, which should, in turn, enable more relevant recommendations.

The Big Picture: Cloudy with some clearing

As noted media expert Jeff Jarvis has observed, not all news is new. Previous and related content often contain relevant information, especially for professionals and others interested in patterns, trends and “seeing the big picture.” Yet few publishers produce content that meets this need. Users are left to fend for themselves or subscribe to services such as Recorded Future, which shows related developments in an informative timeline.

Startups are attempting to bring the big picture into focus in several ways. For example, each day Newstap.es, a news concierge started by journalist Marie-Catherine Beuth, takes a story in the news and provides a brief overview, a more in-depth backgrounder and discussion of a specific issue within the larger story. “Wonk blogger” Ezra Klein recently left the Washington Post to create the world’s first “hybrid news site/encyclopedia” – his goal is to “explain the news,” not merely report it. Circa also allows users to follow stories and receive updates as the story develops. By revealing the big picture, these and other solutions can help users close part of the content discovery gap.

Social: Leveraging the signals in your social web

Social media play a dual role in the Content Discovery Gap. On the one hand, Twitter, Facebook, LinkedIn and other social media platforms introduce their audiences to new content and sources. However, these ever-expanding streams often overwhelm. Each platform recommends certain content, but the relevance varies widely and is often incredibly poor. Gawker, Buzzfeed, Outbrain and others add to the noise with tantalizing headlines and links to content that, like snack food, might satisfy a craving, but lack substance.

Though social media contributes to the volume and noise, it also holds the key to closing the gap. Who we follow on social media and the links we read reveal a great deal about our interests. Leveraging these signals, News.me provides an email digest of “the most interesting news flowing through your Twitter stream.” Little Bird (founded by former tech journalist Marshall Kirkpatrick) uses social media to help users interested in a particular topic identify leading experts and influencers. On content discovery platform Pugmarks, links that are recommended by members of your social network are given higher priority, based on the premise that those individuals’ opinions are more useful to you.

Personalization: The Holy Grail

Companies such as Amazon, Spotify and Netflix use machine learning and other algorithmic tools to recommend selections that closely match a user’s preferences. Leveraged effectively, personalization takes much of the guesswork out of content discovery – on Netflix, for example, 75 percent of viewers’ activity is driven by recommendations. The equivalent for news – what Jeff Jarvis call his News Pal – has not yet been built for two reasons. First, building a recommendation system requires a lot of data about the user’s preferences. Unlike books, music and movies, readers gather news from multiple sources, so data on one’s interests are scattered across websites. Secondly, building and tweaking recommendation systems is a massive effort, requiring highly skilled data scientists. Few news organizations have made the investment.

Startups and publishers are addressing these problems in different ways. Zite (acquired by CNN in 2012, but recently spun out to join Flipboard) and Gravity (acquired by AOL in 2013) have signed up publishers and track users’ reading patterns across participating sites to compile a more robust “interest graph” for a reader. When an individual arrives at a publisher’s site, the interest graph is used to personalize the content presented. Taking a different approach, Pugmarks’ opt-in “trusted companion” accompanies subscribers as they browse and read stories – while reading a particular story, users are shown recommended content prioritized based in part on their own social networks.

Jeff Bezos’ acquisition of the Washington Post may also harken a new era, with Amazon-like technology deployed more widely by publishers. While it may seem like the Holy Grail, content tailored to users’ needs and interests is within reach.

Dr. Phil Hendrix is head of immr and an analyst for Gigaom Research. He is an advisor to Pugmarks.