Home » Social media marketing » Revisiting the question of online influence

Revisiting the question of online influence

I have acute déjà vu. I have just read about a firm targeting solely social media users with Klout scores of 60 and above as potential investors. This raises the whole question of social media and online influence once again, so I think it’s time I revisited it on this blog.

I last ranted posted about Klout, PeerIndex et al in early 2011 after meeting social media writer Brian Solis and articulated my scepticism of online influence scores as reflecting reality, especially given the credence that many employers were putting on such markers at interview. I was immediately shot down in a post by a keen advocate of PeerIndex (citing a tweet of mine) that people only seem to oppose social media influence scores if theirs is not particularly high – i.e. sour grapes. This is not the case and I’ll give you a case in point as to why.

Social media influence in perspective

When I went back agency-side for a year to co-lead the worldwide roll-out of a global blue chip electronics company’s social media engagement platform my Klout score was more than 50 in the old algorithm, which was then a good score. Now 60 is apparently the equivalent benchmark.

After a year of extremely busy work not just working on agency clients and new business worldwide but also the UK side, I did not tweet as often as previously and my personal blogging dried up. My Klout score was just 33 by the time I returned to freelance content marketing in May 2012. In one metric’s eyes I was therefore not ‘influential’ online.

But wait! During that busy time when my Klout score was falling and I was working agency-side – in the real world beyond my personal Twittersphere – thousands of people worldwide were watching the videos the team had created, were liking the Facebook page, reading our posts, enjoying the apps. I was a cog in this content machine but was part of creating genuine online influence on perception and behaviour, not just gaining a number according to an algorithm because I tweeted a lot about football or something similarly irrelevant to my day job. These elements cannot be factored into online influence metrics, which rely on a user’s personal Twitter or Facebook activity.

Social media influence metrics do have some value

I’m not totally dissing social media influence metrics, they do have some value as part of the wider perspective and context, but employers should not look at a score and dismiss that individual out of hand. Had I applied for an agency job (rather than return to freelancing) with a Klout score of 33 and come across someone who deemed that a major factor in the selection process, another candidate with potentially less experience could have got the job based purely on an algorithm made by a machine, which itself can be influenced, of course.

In short, as with everything, look at the bigger picture when gauging online influence. Here’s another related post on how to measure online influence.

What do you think about online influence measurement? What should we be looking out for in 2013?

Tags: , , , ,

6 Responses to Revisiting the question of online influence

  1. Great post Chris. I think the value of influencer scores is that they prove you know what you’re talking about online and that is valuable because many consultants are offering online services knowing people want social skills without necessarily ‘being there’ themselves. But they also have limits as you rightly point out. Your portfolio would stand up under scrutiny anyway. Continues to remind me that social is not REPLACING face to face – it’s a part of it. Thanks for a great post.

  2. Thanks for your comments, Dionne. It is a question I get asked a lot by people and I think that any employer who WOULD regard a Klout/PeerIndex/Kred score over the bigger picture is frankly showing their own ignorance of digital marketing.

  3. I manage multiple Twitter accounts and was curious to see what the “influence” raters would think of them. I have a few accounts that are 100% automated. No interactive tweets of any kind (at the request of the client). It was very little surprise to me that the fully-automated accounts continue to climb on Klout while staying more steady on PeerIndex and Kred. Ironically, my fully interactive clients have seen vastly wavering returns from Klout while climbing on PeerIndex and Kred.

    I understand that what I did is not necessarily indicative of what everyone will see, but found it very curious.

    Thanks for the post. It’s a great addition to the dialogue about influence metrics.

  4. Thanks Stan, that’s really interesting. Have you posted about the findings? I’m really over the whole culture of judgement by points. It’s something some people need more than others: validation by numbers.

    This topic will run and run…

    Thanks again for your comment.

  5. Craig Hughes

    Hi Chris

    Thanks for revisiting the topic.

    I agree that, in the context of something as complex as hiring, influence scores should not be applied without qualitative, personal judgement being used. Filtering applicants out based on their influence score, before any other assessment criteria is applied, might not be wise. No one metric alone should be used to judge an applicant.

    As much as we feel uncomfortable with them, there will always be hiring heuristics used – such as filtering out applicants who didn’t attend prestigious Universities/Colleges – because they are indicative of something and are therefore useful on some level (apologies for being vague – it’s a topic unto itself!).

    In the example of hiring a Marketing/PR/Social Media related role, you might argue that taking into account an influence metric would be wise as it assesses an individual’s ability to consistently drive conversation and engagement online. Perhaps it should be considered, alongside examples of previous successes for clients/employers, growth in online brand presence etc?

    You touch upon something important with your point about reduced personal presence in favour of other online activity (clients/new business) having an impact on your score, when in fact your online influence could be said to have increased as a result of that work. Currently, your work in that respect is not incorporated into your influence score in any meaningful way (although all of us in the space are working hard on solutions to that problem). The same goes for offline influence.

    We (PeerIndex) have data scientists and engineers that are working on gathering (within reason), understanding and incorporating every data point imaginable to allow us to continue to improve our measurement. You can see a recent post from Ferenc Huszar, on the Rising Science of Social Influence here:


    2013 is going to be an important year for the science of influence measurement and the marketing industry’s use of this new data source. Marketers are going to benefit from the addition of influence to their segmentation toolset, an enhanced ability to understand their most valuable advocates and an opportunity to discover and engage the people that their target audience listen to online.

    We have long understood the value of engaging influencers (celeb endorsement to blogger outreach) and an element of measurement has (or at least should have!) been used when determining who to engage. For blogger outreach, how many views do their posts receive? How many comments/clicks/likes/shares etc do they receive when they post? What is the demographic make up of their audience? Tools like PeerIndex, Klout and Kred are now giving marketers the ability to know this value at scale, for millions of people globally.

    No doubt there is an awful lot of work to do to guide marketers in how best to use this (that is a key part of my job incidentally). As with any marketing tool, some will do it well and some will do it poorly. Our experience is that engagement campaigns via PeerIndex etc are a brilliant way to create spikes in engagement, positive conversation and awareness but these campaigns are just one way of using influencer data. Long term engagement plans should be built with influence metrics in mind.

    It would be great to get your (and your readers’) thoughts on how you think influence data could/should be used? Have you seen campaigns that you thought were effective?

    I’d be more than happy to answer any questions you may have.


  6. Thanks for taking the time to comment in such detail, Craig. I’m not *entirely* dissing metrics, I have used them before as part of the entire blogger influence when discerning which bloggers we should target for client product reviews. However, I do take figures with a huge kilo of salt.

    Social media monickers within organisations can rely on (or decree from above) multiple retweets from staff, for example, which helps. Freelancers like me can’t depend on that and even on days when my posts have been tweeted 115 times, as was recently the case, my scores were barely impacted. (I checked as part of my research for this post).

    I’m sure PeerIndex and its ilk will improve as we go forward, which I welcome.

    Thanks again for your comment.