Influence on the Social Web
Our influence algorithm was created to provide an effective way to rank social results, and its utility has expanded over the years as businesses have discovered other uses for influence. Conceptually, our influence algorithm ranks social content based upon people citations, similar to how traditional search engines use inbound link citations to rank pages (like Google PageRank), with people’s influence calculated based upon the level of attention they are likely to generate from the content they post.
Details on influence and how our algorithm works:
- Why am I influential?
- Topsy Influence measures the likelihood that, each time you say something, people will pay attention. Influence for Twitter users is computed using all historical retweets: millons of real, public statements indicating who’s listening to whom. On our website, roughly the top 0.2% most influential of all Twitter users are tagged “Highly Influential”, and “Influential” tags appear for the top 0.5% most influential Twitter users. So if Topsy says you are influential, you are part of a pretty small group!
- Influence matters
- Topsy search results are ranked by how often they are cited in tweets, and how influential the people are who tweet about them. We use influence for features across the site such as identifying Expertsor the best Twitter Profile for a search term, for discovering trending terms and authors associated with a website. Influence is key to many of Topsy’s existing and upcoming features and it is available to developers through Topsy’s Otter API.
- Influence is dynamic
- Influence is dynamic across multiple dimensions and is not as simple as segmenting people within static categories who tweet a lot – saying a lot more is not, on its own, enough to increase your influence. Your Topsy Influence increases when you say things other people find interesting. This means influence changes over time, and what you are influential about – by keyword, category, domain, etc – will change based upon what others find interesting. And, retweets are transitive. If Alice retweets a tweet from Bob, and Carol retweets Alice, Bob’s tweet has not only reached his and Alice’s followers, but also Carol’s. Influence is also transitive – the more influential people retweet you, the higher your influence will be.
- Influence is Like Wealth
- Influence is like wealth – you earn it. But you also spend a bit when you say things that draw no attention from people. Like wealth, your influence increases when you earn more than you spend. Topsy adjusts influence in relation to the entire “economy” of attention, making it difficult for individuals or groups to artificially inflate their influence.
- We recognize that while search results change in realtime, peoples’ influence is relatively stable, built up over long periods of time. As in the real world, influence on Topsy does not change very fast but it does change. We track spending in realtime and earnings at the end of every month. For a more technical explanation on how Topsy processes Twitter data read this blog post.