The Velocity of Ideas and the Information Economy
Content tends to take on a power law distribution = “long tail”. As more people rely on social graphs for link sharing and discovery, and as the velocity of information they are sharing increases, a lot of good content is drowned out by the ‘top’ content – in terms of the 80/20 rule which governs the long tail distribution. What this means is that about half of what’s published hardly gets any page views.
I believe that human nature reinforces a ‘herd mentality’ around the same content, owners, and sites. Unique discovery is repressed by the rabid sharing that goes on at the “fat” part of the content tail (the content with a high ‘velocity’).
Why? Because the feedback loop goes as follows—‘hot’ content is shared among peers, and people often retweet or share stuff which makes them appear informed (“look who I follow”, “look what I know”). Anyone who uses Twitter can see this issue. People are hesitant about sharing something that isn’t already ‘validated’ by an element of virality, even if they believe the content is of value.
Velocity of Information in Relation to Abundance:
At SXSW, Clay Shirky gave a talk about information abundance / scarcity. This has been a popular topic over the last few years. Clay argued that abundance is a bigger problem to society than scarcity, since when something is scarce, it’s easy to handle by simply increasing the price of the item that is scarce. When something is in overabundance, people have trouble pricing and valuing it.
I agree and would argue that in the case of content, a relative overabundance coupled with the ease at which content can be shared causes multiple issues:
- Value assessment: Overabundance of content gives people a reduced propensity (incentive?) to ‘spend’ time curating, and the value of unique ideas falls in terms of their velocity and reach… another way of saying this is that the high velocity content drowns out the rest of what’s out there.
- Influence: people are over-influenced by their peers to read/share popular content—e.g. “100 people tweeted this, it must be worth reading”. This causes a person to fear that if she doesn’t consume the same info, she will be ‘left behind’ or uninformed. In this way high-velocity content consumes the person’s limited attention span (attention is scarce).
To be clear, I think social recommendation and discovery are incredible, but signal to noise problems and issues related to human incentives limit their utility. This works as a force inhibiting discovery of truly great content. Because of these human limitations to discovery and matching, I think machine learning and matching technologies have the capability to best crowdsourcing and social sharing.
Hopefully this area will see continued innovation and VC dollars, as companies such as Hunch and LiveIntent use technology to revolutionize discovery and recommendations. Feel free to use your favorite aggregation site or Twitter to share or debate these ideas