Advertising exists to seed memes, ideas that spread through society like viruses in your head on a cold winter’s day. The fundamental hope is that a product or service concept becomes so desirable that, like Razor Scooters in the late 1990s or glass-tablet gadgets today, suddenly everyone will want one and share the news with their neighbors, with no incremental marketing cost. This is why the lure of social media is powerful with its promise of free, scalable connections …
Yet a new whitepaper by Cornell and HP Labs suggests building fame in social media is tougher than you think. Daniel M. Romero and colleagues processed more than 22 million tweets from 12 days in September 2009, looked at how often people clicked on web links inside the tweets, and then compared how those people were connected — and found that pure number of followers does not equal influence. Instead, as you might guess, there are some entities in social media with many followers whom no one listens to, and conversely some with few connections who tend to have their ideas shared everywhere. It’s an important dynamic to understand if you try to spread messages in Twitter, because only 1 in 318 tweets with URLs is ever retweeted, meaning the vast majority of Twitter missives trying lure people to click links hit a brick wall.
The Cornell/HP report found four types of entities on Twitter:
2. The passive — users who follow many, but rarely share things (“lurkers” or “automatons” such as @redscarebot).
3. Those not influential with many followers (“all show, no go,” alas, like @newsweek).
4. The highly influential with comparatively few followers (“beacons” such as @twitdraw).
The Economist suggests new analytics services such as Gnip will profit from helping marketers identify the human nodes inside social media that fall in buckets 1 or 4 above. Not everyone is an influencer; in new media, the money may flow to those who act most like old broadcasters.