Presenting Pearls: Stakes of Content Discovery

Users of fast growing services face challenges discovering relevant content. To address these challenges is hard because relevance is an ever evolving concept which depends on the context. When the user has a clear goal or specific question, relevance is straight forward: that’s how we got search engines. But it is less obvious to address the needs of users who want to discover content that they don’t know about yet, or who watch a specific topic over a period of time. Search is about asking questions and getting answers. It doesn’t help you to figure out which questions to ask. recently unveiled their Topic Browser to address these issues and let users of the platform extract more value out of our collective curation efforts.

Problems of Visibility

As the number of users to social networks and publishing platforms grow, the signals multiply. On the one hand, owners of the platform celebrate the growth of their service and the success of their company. On the other, this threatens to diminish the value of the platform as users get overwhelmed. The social signals which were supposed to help them make sense of the tidal waves of content become a part of the waves themselves. Moreover, potential users have difficulties understanding what the service is about or where the best stuff is. Each growing platform faces the same issues: Twitter, Tumblr, and

In our world of plenty, each service wants to give users tools to navigate the oceans of content and find the true pearls. If these tools can be developed, problems of visibility will not only be averted but the community will get more value.

Content Discovery Mechanisms

Relying on content discovery mechanisms means that we are outsourcing part of our choices. They present us with manageable amounts of options and we choose from this instead of choosing from the bigger pool. This is how we get a suggested user list on Twitter, Flickr’s interestingness index,’s homepage, or dedicated services such as Squidoo or and its Topic Browser.
Tumblr faced a problem with their vast amounts of signals. Tumblr relies on tags and human editors to surface the best content. Tag pages are produced by contributors and editors working together. Users, then, can subscribe to this curated experience.

Squidoo is another service which relies heavily on human editorial skills to create pages about topics with original content and material from around the web: images, videos, RSS feeds. The addition of Amazon and Ebay affiliate modules brings revenue that is shared with editors who can then give it to charity or keep it. is in a position to solve this problem differently. With their Topic Browser, they inverted the process. Instead of having topics curated by few editors, they rely on the massive numbers of Paper.lis edited every day by their algorithms in tight collaboration with their users to surface content. The scale at which the Topic Browser operates is impressive: more than 13 million articles are categorized in one or more of the 20,000 curated topic pages. The filters ensure that the most topically relevant content gets added. Each topic page also features links to the individual Paper.lis which contribute to it, making it a topic watch as well as a discovery tool.

Personality Through Taxonomy

Using the wisdom of crowds to perform editorial tasks may seem risky but it gives us access to new information about our own community. On the one hand, individual personalities of the community’s members and publications may not shine as much anymore. This is a loss, since personalities are to be cherished. Yet, on the other hand, we may discover more about our collective biases and quirks. The culture of services and the communities they gather shine through their taxonomic choices. Tagging and categorization practices give us insights into the choices made when developing algorithms, and about human editioral practices of both administrators and users.

On Tumblr, for example, the Explore function is based on enlisting moderators to curate topic pages. They introduced this function with a limited array of topics in December 2010 and slowly added more to finally merge everything into the Explore function. Today, the page shows a selection of what the service has to offer.

This page does say tons about the Tumblr community: their love of pictures, leisure-related content and cuteness becomes apparent.’s Topic Browser is still in Alpha but the tagging system already shows signs of personality — especially on broad topics. Recreation, for example, seems to be synonymous with exercise for most of our community’s members. Love’s topic page points to an impressive array of posts about interpersonal relationships, wonderful personal blogs, and some NSFW images forming a striking portrait of the topic online.

As we explore the Topic Browser together, we will gain new insights about our community and our world. Dive in. And when you come back out, tell us what pearls you found and what you learned about the community in the process.

Image credit: “Pearls” by Dr. John Supan for the U.S. National Oceanic and Atmospheric Administration.

I wrote « Presenting Pearls: Stakes of Content Discovery » on’s blog, it was originally published on February 6, 2012. Reproduced here with permission.





Étiquettes :