Editor’s Note: The article has been reproduced from Platform Thinking written by Sangeet Paul Choudary. He is a widely published technology analyst, startup advisor and innovation researcher. He collaborates with economists at the MIT Centre for Digital Business on his research on internet business models and is a mentor at leading accelerators like 500 Startups, JFDI and Founders Institute.
You can’t invite people to a party and try to sell them stuff. Pretty much every starry-eyed startup that went after eyeballs gets it by now. Over the last seven years the web has moved away from a consumption medium (think NY times) to a creation-consumption medium (think Twitter, Facebook). But we’ve been very tardy in reshaping business models for this new model of the web. Interestingly, the solution to this monetization problem may lie with a small insignificant key on your keyboard. Read on.
Why are we failing at monetization today? Traditional online media worked on a Pipe model, targeted only consumers and got away with monetizing eyeballs. Social media works on the Platform model, supports both creators and consumers, and has tellingly failed with trying the same old monetization strategies.
Media Monetization 101
The monetization of any form of media is driven by mining of context and using that (or some other consumer action) as a proxy for intent. Advertisers then pay to have their ads matched with the right intent. Here are a few examples:
Keywords on a page: Context E.g. AdSense
Search query: Intent E.g. AdWords
Location: Context E.g. FourSquare
Monetization works by harvesting user intent and serving messages/information relevant to that intent. The better you are at harvesting intent, the more effective your monetization is going to be.
Why is this model breaking down?
Mining context and intent goes for a toss in the world of social platforms. Users are the new content creators and content isn’t necessarily structured. With the older media model, the content creators (typically the media houses) were creating content to cater to search engines. The content was designed for text mining algorithms right at the point of production. With social media, the creators of content (all of us) don’t care about structure. In fact, online conversations are getting more unstructured by the day. Consequently, mining these conversations for context and intent is a crazy task, riddled with false positives. And false positives always lead to spam.
This is why the Hashtag is so important to the future of the web.
Enter the Hashtag
Engineers would like to be known for the tech innovations that they engineered but Chris Messina will probably go down in history as the guy whose random blog post helped structure a new era of media. In a 2007 post, Messina suggested the use of Hashtags for the first time for Twitter.
This week, Facebook rolled out Hashtags.
Hashtags and Platform Thinking
If you think of media as a Pipe where content creators create stuff and push it out for us to consume, the content creator takes great pains to structure the content. Every piece of content will be carefully drafted in a category, will be peppered with keywords for search engines to gobble and will be structured so that the context can be easily mined.
If you look at the proposals from Stephanie and Brian, they advocate the use of pre-defined groups to regulate conversations around certain contexts. This is a typical Pipe Thinking model. Provide the constraints and force the creators to work within those constraints. It works very well when media is created within the boundaries of a firm.
When media is created by users, as it is today, one cannot afford to think in terms of constraints anymore. This is where Messina’s advocacy of the Hashtag is so brilliant. If you’re thinking in terms of Platforms, you’d want to make the creation process as easy as possible for users, yet ensure that they leave you with enough hints around intent and context. This is what Flickr did when it allowed users to tag pictures instead of forcing them to fit pictures into pre-defined categories. This is what Messina advocates in this post when he argues against users having to operate within groups and allows users to define context and intent on the fly.
Top-down classification and forcing creators to fit within categories or groups is a hangover from Pipe Thinking; an editorial view of the web. A social view of the web requires a more bottom-up approach.
If you think of the social web as a flow of information, pre-defined categories and groups limit the channels in which information can flow. Hashtags, instead, allow creation of channels on the fly to suit the needs of the information creator.
If you’re still thinking Semantic search alone, you’re in the wrong game
When the world first saw an explosion of user-generated content, people realized that Google’s keyword and link-driven approach to ranking information wasn’t going to work forever. Semantic search was hailed as the next savior.
I have nothing against semantic search. I just believe algorithms are still fairly limited in mining human intent from unstructured conversations. And the web is gradually, but definitively, moving towards unstructured conversations.
The solution to mining unstructured information doesn’t lie in creation of more sophisticated algorithms alone. It lies in, first, solving the problem at the point of production and allowing the new creators to easily append some structure to the information.
That is exactly what the Hashtag does!
If you’re building a platform that enables and promotes unstructured conversations, and you want to go beyond just being a communication tool, to creating a corpus of sticky content, hashtags can help transform unstructured conversations to structure, right at the source.