Klout: influence is more than a numbers game
Real life influence is far more complex than a simple metric - as the debate around a list of advertising and marketing professionals has spotlighted. Brand Republic published a list of the most influential people in advertising and marketing on social media in the UK on Monday.
It asked Klout to run the numbers on influencers on 15 advertising and marketing related topics over a 90-day period.
Klout is a website and mobile app that uses social media analysis to rank its users according to online social influence via a so-called Klout Score, a numerical value between 1 and 100.
In a Twitter chat using the unashamedly hyped-up hashtag #UKTweetElite, several of the people ranked in the list celebrated their new found status and shared tips for social media engagement.
We all love a list. They appeal to our ego and competitive nature. As a result they're incredibly social.
In the days that followed the list was noisily debated on Twitter with criticism levelled at Klout for its lack of transparency and crude, simplistic approach to influencer identification and social media influence.
By Brand Republic’s own admission Klout is controversial. The project generated much discussion and more than 1,700 shares.
Social media listening tool Lissted produced a rival list based on analysing 15 million Twitter relationships.
UK Twitter boss Bruce Daisley said that the Lissted list “looked much more accurate.”
I've made a list of my own. I asked 13 executives, planners and strategists in my network how they rated Klout as a tool for identifying and measuring influence.
Many of these individuals, like me, were named in the Brand Republic or Lissted lists.
Their responses call out Klout as a black box that is easily-gamed.
Influence can’t be simplified to a single number. We all view influence from a personal perspective.
Measuring influence in social networks requires listening as well as an understanding of context and relationships - real life influence is much more complex.
#1 Klout isn’t social
We don’t know what it measures, how it measures and what score or rating they give to different variables. In my view this makes the results meaningless.
Not everyone agrees with me. In fact there are a number of clubs that game the Klout system. Sharing and favouriting each other’s posts and tweets to raise their scores. This misses the point of social.
I love sharing posts that I know my audience might like. I want to have a relationship, engage and have a conversation. It gives me a Klout score for sure. Maybe not as high as some, but I don’t care.
I care what my connected communities think. That is social and it’s how we influence and are influenced.
#2 Activity doesn’t equate to influence
It’s closer to an attempt to use social data to assess someone’s potential to influence. Unfortunately its approach appears far too simplistic to be fit for this purpose as it still looks to be primarily a social media noise measure, seemingly ranking accounts higher on factors like reach, volume of tweets and mentions.
In a social media context, the ability to manufacture noise and game the system means that such measures are dubious.
Also activity on social platforms is just one way that someone could seek to influence and in many cases it won’t be the primary one.
To address complex questions like who might be influential you need to delve deeper to look for more complex social signals, like the significance and relevance of someone’s relationships and interactions in a particular community.
#3 Reward for frivolity and spam
Based on that I should probably jack in the sometimes intellectually-draining pursuit of writing articles for my blog as well as guest posts and embrace the cult of the frivolous.
Posting only ‘like-bait’ photos of my kids on Facebook instead. And ensuring every Tweet I write includes the hashtag #contentmarketing as these tweets are favourited and retweeted automatically by several corporate and personal Twitter handles – so swelling my Klout score.
Klout doesn’t work because it’s easily gamed and includes a broad-based measure on seven social media networks which we all use differently to participate in conversations with different factions of our lives.
#4 Scores are snapshot of activity not a measure of influence
For example, according to my Klout profile, there are lots of areas it rates me as being an expert or having influence in, such as: blogging, journalism and social networks.
However, it also includes the BBC and Manchester as some of my key topics. Well I certainly enjoy watching, reading and listening to what the BBC produce, but being an expert?
Social influence is a growing marketplace and looks set to stay. I place importance on online reputation, including how you manage it, rather than getting hung up on Klout scores or similar numbers.
It’s a snapshot view of how someone interacts within the communities they’re part of, but isn’t failsafe or an accurate reflection.
#5 Influence is relative
Former Klout-rival Peerindex attempted to create a topic relative social score but has dropped this aspect of its service [after it was acquired by BrandWatch in December 2014].
Having said all this, it is hard to avoid Klout scores. Countless social tools such as Hootsuite incorporate it. The biggest challenge is that Klout gives people what they want, even if they know it probably isn’t right, namely, a single number by which to rate people comparatively.
The safest option at the moment is to use Klout sparingly and in specific use cases. But you certainly wouldn’t want to base your entire influencer approach solely on a Klout score.
Better still, consider alternative approaches which attempt to go beyond obvious output metrics and understand relationships and impact relative to specific communities.
#6 Users reject personal profiles
I’m ranked as an expert in topics including advertising, b2b, community management, marketing, Nokia, social media and Weiden+Kennedy.
I’d say that I don’t love a single one of these topics, so the purpose immediately jars with me.
Klout is purely pulling my tweets and looking at the number of times I mention something, probably combined with reach/response to those tweets.
And then, Nokia - I haven’t talked about them for years – so they aren’t even taking recency into consideration.
It makes sense from a user perspective that they’d introduce recommended content from the most influential sources, apart from the fact that it’s not. It’s just a random stream of content that apparently relates to my interests, and for the most part, it’s a bit rubbish.
#7 As relevant as a unicorn
We use Brandwatch and Sysomos for influencer identification as they have more sophisticated algorithms that scrutinise relationships.
Klout fails for me in being able to give contextual influence - it told me I was influential in unicorns once and I don't think I've looked at it since.
#8 Vanity tool that doesn’t relate to true influence
Colour me cynical, but I see Klout and most other influence measurement programs as vanity tools. But do they have impact? Perhaps if you make it onto a Klout-based "influencer" list you will get some free stuff or a new client or a speaking gig; however, unless you actually do superior work and deliver influential expertise (that is listened to), the impact of this mark of distinction won't matter for very long. But hey, go ahead and put it on your Twitter bio.
When I heard Arianna Huffington speak last November she told the enraptured audience that the true impact we have will be revealed in the eulogies those closest to us will write and say upon our death. Never, she asserted, has how a company performed in any financial quarter under someone's watch been a part of the remembrances of anyone of true significance. Rather it was how the person made a difference in lives, whether for the greater good of society or immediate family and friends.
#9 Reputation scores aren’t social
Klout is a reputation scoring merchants. If you let them access your Twitter, Facebook and other social media profiles they will give you an influence score. On one level this just formalises one of the favourite games of social media early adopters.
Gamefication begets gaming, or at least more overt gaming.
I think that sometimes successful players are distorting their reputation and personal networks - maybe that’s not that useful in the long run.
Online reputation is an extension, a real part of your actual reputation, not a game world. Metrics and incentives can be very useful, as any manager worth their salt will tell you. They can also warp behaviours and destroy value inadvertently.
Behaving differently to boost a score is a slippery slope that may be corrosive to your real reputation even as the numbers climb higher.
#10 Single scores miss context and relevance
The thing about influencer scores is that they don't provide sufficient context or relevance, which is why so few marketing professionals or executives trust them.
I have a few followers and many of them have been acquired as a result of campaigns or media testing.
I tell some good stories but I reckon only 5,000 of the 72,000 are of any real value. If 10 people RT or fav something that's a big deal to me.
Kred was close to getting it right as they included influence and outreach. Big difference.
#11 Relationships are the key to influence
Suddenly I was an expert in “water” and friends become experts in lawn darts, golf, and liquor.
Then you saw the shift to employer’s looking at candidate’s Klout scores to determine if they were worthy of jobs and all hell broke loose.
Today Klout is too easy to game and is reliant on the wrong things.
There are plenty of people with Klout scores in the 40s who are extremely influential in their circles, but an organization would never get them a second look.
Let’s focus on what really matters and less on the vanity metrics.
#12 Influence isn’t a single number
Whether it is the impact of offline versus online influence or the subtleties of reach, relevance and resonance, I’ve yet to find any fully automated process that even comes close to giving a comprehensive answer to the problem of influencer identification or measurement.
Technology has a role to play but, in my experience, it can never be used as a short-cut to real insight.
#13 Networks tell the true story of influence
Understanding influence and influencers is not a straightforward process. It’s always going to be open to abuse if we reply on mass algorithms on SAAS platforms working on a one size fits all approach.
Influence is a real world thing. The software approach looking at all of the content noise out there and working back in will continue to make mistakes – wrongly assigning influence to the noisy people, missing people who are quiet or non-existent online but carry massive influence in reality, and miscategorising people and their topics of relevance.
Klout has some benefits and merits in the same way that it can be argued that the much criticised AVEs (advertising value equivalents) do. They both provide indexes that go up and down, but both are flawed and neither actually offer what they at first appear to claim. Understanding influence requires better tools, factoring in real world characteristics, desk research and a great understanding of network connections and relationships.