Many tools like Klout, Tweetlevel and Peerindex are available online to derive the influence of a user/brand on social media platforms like Twitter and Facebook. Such tools use algorithms to find out the number of followers and interactions, based on which it concludes the online influence of a user/brand. But, is it the best way to trust algorithms in order to figure out the online influence of a user/brand, or should there be an alternative?
Head, digital strategy, Edelman
Online influence should be calculated using a combination of algorithms and human intelligence.
Algorithms help in narrowing the range and scope of people that can be added into the ambit of influence (for any specific sector or vertical), while human intelligence can help validate the algorithm in terms of relevance. This dual-level check helps to ensure that brands identify and target the right influencers, and do not seem irresponsible and wayward with influencer outreach.
In terms of online tools for mining influence, it is better if those tools also reveal the kind of networks influencers most often use to influence.
Broader influence-ranking tools may end up with a single score from across multiple tools, but that may not explain the nature of usage of each tool, by each influencer. Some may use Facebook, for instance, purely for personal communication, while using LinkedIn and Twitter for broader, professional communications. Tool-level influence ranking tools will be able to make this distinction and be more accurate for brand outreach.
Senior knowledge manager, Blogworks
There are several tools that give an automated influence score, but their accuracy is still a question mark for several reasons.
First, most of these tools are limited to specific channels and don't take into account the entire online presence. For instance, a person may be influential on a blog, but if the tool does not track it, the influence will not be accurate in that case.
Second, 'high popularity does not imply high influence and vice-versa'. While many tools are trying to delve beyond populist metrics such as the number of followers, they are quite vulnerable to manipulation. Third, while standardisation is important, a one-size-fits-all approach does not work well.
An effective tool should work on a series of metrics (some may find topic posts to be of equal, or more value, than just number of visitors and inbound link count).
The answer lies in finding the balance between automation and human intervention. Automation can scale up influencer identification process. Human intelligence can keep a tab on the accuracy and relevance of results to the domain under consideration.
Director, digital and social media, MSLGROUP Asia
Influence is difficult to measure in general, and online influence is difficult to measure in particular. Influencers get their credibility from different sources; their name recognition, expertise, reputation, network, and inside knowledge.
Tools like Klout and Peerindex only measure the strength of the influencers' networks, not the other sources of influence. The CMO of a Fortune global 500 firm will score high on several sources of influence, including the strength of her professional network. However, she may have a small online network, and therefore, a low influence score according to a Klout or a Peerindex.
The tools mentioned earlier are most effective in identifying grassroots influencers. For such influencers, the other sources of influence -- name recognition, expertise, reputation, offline network, and inside knowledge -- often lead to a strong online network and a high Klout and Peerindex score.
Increasingly, an influencer's strong online network will also strengthen her other sources of credibility and increase her overall influence.