This is part of a series of posts based on excerpts from The Realtime Report’s Guide to Influence Measurement Tools.
It’s hard to know exactly what an influence score measures. But can we tell if a given user’s score on a platform like Klout, Kred, PeerIndex or TweetLevel is a “good” score?
All of the people-scoring tools assign a single overall score to each user. Some create additional scores around a user’s online influence within certain topics, or for specific attributes. But how many users have scores at different levels? How can you tell if a score is high, low or somewhere in between? Unfortunately, most of the personal influence measurement tools are not willing to share details about how the scores are distributed.
But without knowing the distribution of the scores, there is no way to understand how to interpret them, except in context to other scores. You can’t look at an individual user’s score and say “this is a high score.” You can only look at a given set of users and say “this user has a higher score than that one.”
Only Kred and PeerIndex were willing to share some statistics about the score distribution on their platforms. Klout simply says that its average score is “around 20.”
Klout assigns only one general score to each user, but some of the tools offer additional metrics. TweetLevel displays individual scores for popularity, engagement and trust. Kred also assigns a score for outreach—how actively the user engages with other members of the community. PeerIndex and Kred both look at a user’s activity in the context of specific topical communities, so their scores are calculated relative to other members of a given community.
Here’s how the scores work, and what each of the personal influence measurement tools are willing to share about the distribution of the scores on their platforms:
Klout: According to Klout’s former Senior Marketing Manager Megan Berry, a score of 50+ (out of a possible 100) is in the 95th percentile. The average score is “around 20.”
Kred: The “Global Kred” of the most influential user on Kred is set at 1000. Kred also reports a score for “Outreach Level,” which measures the user’s level of activity within a community. Scores are also assigned in each of
9 70 topic-based communities. A score of 600+ puts a user in the top 21%; a score of 800+ would be top 0.1%. The average score is not published for “Global Kred,” but Kred does show the average scores for each “Community” on the Community pages. UPDATE: Kred does publish its score distribution at http://kred.com/rules (click on Influence in the left tool bar).
PeerIndex: Until recently, PeerIndex reported both an overall influence score (based on a scale of 100), as well as separate scores for “Activity,” “Audience” and ‘Authority.” As of March 2012, those more nuanced scores no longer appear on user profiles, although they are still available via the API. A score of 40+ puts a user in the top 10%; 90+ would be the top 0.1%. PeerIndex does not report an average score.
TweetLevel: Influence scores range up to 100; TweetLevel also reports scores for “Popularity,” “Engagement” and “Trust.” According to TweetLevel creator Jonny Bentwood, “anyone with a score above 65 has done very well”–but he was not willing to share any additional details about the score distribution or the average.
The bottom line: because the personal influence measurement tools provide very little data about the distribution of their scores, looking at individual scores without context is meaningless. The numbers are useful only if analyzed in the context of how different users compare to each other.
Klout scores have often been compared to FICO consumer credit scores, an analogy often used by Klout CEO Joe Fernandez. But there’s an enormous difference: FICO clearly outlines its methodology for arriving at the score. And the distribution of FICO scores is widely available, so consumers and lenders know exactly how to interpret a given score. And all reputable lenders look at three different sources (Experian, Equifax and TransUnion) when evaluating a customer’s credit.
That is not the case with today’s personal influence scoring tools. Klout and its competitors do not provide details on the factors that make up their scores, nor do they publish the distribution of the scores. That’s why it’s critical that anyone using personal influence scores to make business decisions should, just like lenders do, use more than one data source to do so.
To learn more about influence measurement tools, check out The Realtime Report’s Guide to Influence Measurement Tools, our detailed analysis of personal and contextual influence measurement tools.