We have moved! Mango Information Systems (Belgium) is now Alef (Spain).

Finding twitter influencers

the recurring problem of defining influence

This is the first of 2 articles in a series about social media and influence. We review the key concepts and issues in this article, while the second dives into a comparison of social media influence systems.

Meet the controversial measure

Measuring the impact a person’s activity on social media is a recurring topic around the digital marketing world. Klout leads the market for measuring the social influence of individuals. There has been a lot of buzz and hype around it, also controversy around its management of personal data. In general, the concept of measuring “influence” is not well understood by the public, and Klout’s opacity around its methodology has not helped in clarifying how the score is deduced. Klout’s competitors like Kred, differentiate themselves by being more transparent about their methodologies. We will compare some players in the second article of the series.

The widespread use of the term “influence” when talking about social media has led to many misunderstandings about what it actually means. Klout score, or scores provided by Klout’s competitors such peerIndex or Kred, actually measure the “potential for influence”, i.e. how likely it is that the person’s messages are interesting and listened to and therefore shared. This does not necessarily mean that they have an effect on the state of mind reader, which is the notion conveyed by the term “influence”.

An influence score tells us much more than the number of followers. The scores put emphasis on the reactions (shares, comments, favorites) by followers/fans to the author’s posts. The result is a crowdsourcing of the measure of the content’s interest. A user with many followers that posts a lot of uninteresting content will end up with a lower score than a specialist with lower audience, but posting high-quality content.

For more details in how influence measurement works, read this article: Can Social Influence Be Distilled Into A Score? Part I, The Potential.

side illustration: logos of major social media influence measuring systems

How not to use influence scores

Common sense should prevail in the interpretation of Klout scores. They do not represent how important or how valuable an individual is, either on social media, or in real life.

Some businesses use Klout scores to decide what special offers they give to customers. This does not look good and can back-fire, as it might generate frustration from some clients.

On the other hand, a great example of a responsible decision regarding Klout scores is at Mobile Vikings, the leading Belgian leading Mobile Internet provider. I recently had the chance to interview Dorien Aerts, Mobile Viking’s Chief Marketing Officer. She told me:

“All our Vikings (customers, Ed.) are equal, and treated equally. This is why we do not take the influence score into account in our customer support software”

Mobile Vikings refused to use the feature pushing influencer’s support requests to the top of the stack, available in the solution they use. We find this policy very wise and well thought through. A company should ensure the best level of customer satisfaction whomever the customer. We cannot predict the impact of any customer’s satisfaction or dis-satisfaction on the business.

On the other hand, knowing who is influential in a community is a precious piece of intelligence for any marketer willing to expand the reach of their brand.

Why you should care about social media influence

As a marketer, social media influencers connected to your clients are key people. You need to identify who has significant reach within your target markets, and engage in a relationship with these people. You will have to evaluate profiles on a case-by-case basis and act to make the right influencers become your brand ambassadors.

At Mango Information Systems we think that influence measuring systems are too generic, even when they are broken-down by topic. A marketer needs to identify influencers amongst each of their customer segments, and measure influence inside their target market, which is narrower than the analysis done by Klout et al.

Influence measures only take interactions into account, not connections between people.

A person’s position in the social graph (set of connections between persons) is crucial in the spreading of information, and there is always a small set of people who are key in the spreading of a message. Viral marketing is becoming a science, read US Military Scientists Solve the Fundamental Problem of Viral Marketing.

This type of meaningful, actionable identification is exactly what we are doing at Tribalytics, a tool designed to help marketers strategically plan their campaigns on Twitter.

To see how we solved the problem of different influence rankings for a case we worked on with Finn.be, read Part 2: Comparing Klout, Kred and peerIndex.

We have moved! Mango Information Systems (Belgium) is now Alef (Spain).