Scoring social media influence - what’s the story?


Social media influence scoring by KredStoryPeoplebrowsr’s Kred, a social influence and analytics service, today announced the launch of KredStory which is “a new way of seeing social influence that is different than anything we – or any social analytics company – have ever done before”. The good news is that this is an attempt to move away from a score automatically assigned to you.

Beware of influence scoring systems

The scores that each system produces differ drastically. Looking at my scores from Kred, Klout and PeerIndex highlights that these systems are rather mystical. Looking at it simplistically, I would argue that the data available to each of these platforms is the same and therefore my scores should be comparable, clearly that is not the case.

So what are my scores for each?

  • Kred - 639 (out of 1,000)
  • Klout - 51 (out of 100)
  • PeerIndex - 11 (out of 100)

And what does this mean?

To be frank, not much if you are just looking at a score, but before completely writing off these scores it is important to note that I do think they can be of use such as by helping to identify influencers.

Using scoring systems to help identify influencers

For brands that have thousands and thousands of mentions on social media platforms each day, it would be an extremely time intensive (and not to mention costly) process, to review each and all of the people that were responsible. A couple of ways scoring systems can be of use:

1. To narrow down large lists of potential influencers

To make a huge list more manageable you filter to view only the top 10%. Using a tool such as Brandwatch would be a way of doing this as they integrate both Kred and Klout scores. The next step is the most valuable; it is the point where you ‘bring in the humans’. Tools such as Klout and Kred are largely reliant on the number of followers or friends that someone has to determine a score. What they cannot reliably do is tell you of those followers who will be relevant to your client, and who are even real followers and not automated bots.

2. Identify influencers on certain topics

Another aspect which the tools share is the ability to identify topics that people are influential on. For example, Klout tells me that I am influential about football, gym, steak, rugby and burritos (amongst other things). But these are vague and whilst fairly accurate this information on its own is useless. I may be influential about the gym, but of my followers, I know for a fact that only a small proportion are interested in this. So using a tool such as Klout to draw up a list of influencers on a certain topic is only useful if you undertake a thorough research into who the followers are and whether they are a relevant audience for you or your brand.

Does KredStory set it apart from its competitors?

KredStory provides information in a more intuitive and visual manner for people to see what is happening amongst their followers. For brands, it will easily allow them to identify influential individuals around specific topics, and in a more visual manner. In summary, while I don’t think this is a game changer, it may mean that when people are considering using a influence measurement tool, that Kred gets the nod.

What it doesn’t change is the fact that all these tools are merely scratching the tip of the iceberg and that none of these tools can be relied upon without a significant time investment from the people using them and deciphering the information they provide.

Klout coupons for Facebook – will it work?


Audi Facebook content available with Klout score


A new app for Facebook pages will take a user’s Klout score into account before giving them access to certain content.

While exclusivity is great for generating publicity, could this tool risk leaving some legitimate fans feeling snubbed?

Klout measures activity and influence across Twitter and Facebook, using 25+ metrics, and calculates an overall score on a scale of 1-100. The idea behind the Facebook app is that brands can then offer exclusive content, deals and discounts to users who meet a certain influence threshold. Theoretically, this “gating” should reward and capture the attention of social media users who are more likely to share their experience with their audience.

Audi is the first brand to use the technology, and the first “perk” available is a desktop wallpaper - a relatively minor prize but certainly a gesture that I’m sure Audi fans will appreciate.

Klout, like other influence measurement tools, does have some drawbacks - if you’re not satisfied with your rating you can “game” your way to a higher score  so the accuracy of the number may not really reveal much about how much influence a user genuinely has.

Another difficulty is that the quality and areas of interest for your audience are not indicated. Using Audi as an example, even though I have a score of 38, my involvement with automotive discussions and communities is very low and there are equally likely to be petrolhead types who may already be Facebook fans of Audi but don’t have enough of a social network to be considered worthy of additional content, even if they post in specialist car forums.

As a gatekeeper, using Klout risks letting in the wrong kinds of fans, or worse - it could alienate genuine ones. I’m of the opinion that while rewarding loyal and influential social media users will clearly have benefits for word of mouth, tools such as Klout and others may need to become more refined.

That said, maintaining a user’s interest and engagement requires something in return for their time, so I really do appreciate the direction that this is taking. Brands should always be thinking about what they can offer their fans, friends and followers in return for their interest.

Why you shouldn’t join every conversation about your brand online


Keep calm and carry on
Image by scottroberts via Flickr

When brands start social media monitoring, the ability to get real-time alerts whenever your brand is mentioned can be enlightening. Your inbox is suddenly filled, almost in real time, with every mention of your brand. The good, the band, and the ugly. The temptation can be to respond to all of these. To counteract every negative comment. To respond to and then spread every positive experience. To answer and resolve every question. This is only natural for people who care about the brands they work for. But the best approach is often not to respond. In fact, in many if not most instances, a brand should not respond to people talking about it online.

The real benefit of social media monitoring for brands is that it allows you to be aware of and listen in to conversations that you might not have known were going on otherwise. People who express their frustration with your product but would never have told you, advocates telling others just how great you are, or people sharing useful feedback and product development ideas. It’s great to see all of these things and the temptation is to respond. But more often than not, the best thing a brand can do is to not respond.

Doing nothing is often the most difficult thing to do. But it is often the right thing to do. If you overheard two people ranting about your brand on a train you would be unlikely to interrupt. If you heard people talking in a cafe about great customer service they’d received from your team you would probably listen, feel proud and let them tell each other how great you are. There is no need to interrupt in these cases. A rant is probably just a rant and there is little you can do to change this. And people being positive are probably doing lots of good for you on their own without you needing to add anything. Whilst things are different in social media - notably that the comments can be seen by a much larger audience and that they are archived and searchable. But often the same rules apply.

If you have nothing to add, don’t say anything, and if you will only inflame a situation then stay out of it

Overall, brands should be careful about engaging online and have a clear process of when to respond, and when not to respond. There are two very clear cases where a brand should always step in:

  1. Where an actual customer service complaint is being expressed - you should step in to respond to this, pointing people in the direction of where they can get support or dealing with this complaint through your existing channels.
  2. Where incorrect things are being said about your brand, products or organisation - you should correct the incorrect messaging that is being spread and answer any questions

In all other instances you should be more circumspect about getting involved. You should have a simple process for reacting and responding online and use this to help guide you. But overall you should do nothing more than you do something. Monitor, report on and learn from everything people say about you online. But don’t feel the need to get involved in every conversation.

Michelle Obama’s $2.7Bn Influence


michelle obama influencer

Image courtesy of Studio08Denver

There’s been a great deal of talk this week about influence. It’s been driven (dare I say influcenced) by an article in AdAge by Matt Creamer. He takes a swipe at Justin Beiber (dangerous) and points out that automated social media monitoring tools (or influence trackers) like Klout, need some human analysis and insight to get the most out of them.

Clearly there is more to influence than popularity (see this slidshare about online influencers), but popularity can be a pretty key determinant for some influencers. The Oprah Effect is worthy of note as a case in point.

Anyway, all this chatter reminded me of the Michelle Obama Influence Infographic.  The Harvard Business Review recently published reserach by academic David Yermack. He found that there was a strong correlation between the brands Michelle Obama wore and subsequent stock price increases. The percentage increases are small and there is a causation/correlation debate to be had, but when she’s potentially driving $2.7Bn in value for these brands, it’s worthy of note:

influence and michelle obama

Social media influence and other data Twitter doesn’t share


Top secret area
Image by Marcin Wichary via Flickr

Twitter knows the influence of all of its users. But it isn’t yet telling us. This was what we discovered this week at the Web 2.0 summit in San Francisco. When asked a question about how Twitter is able to recommend users so accurately, Twitter co-founder Evan Williams said that they derive the suggestions from a reputation score they calculate for every user. A reputation score Twitter isn’t disclosing yet, but that could be a great tool for finding social media influencers.

This kind of tool would be valuable to everybody who is trying to analyse, understand and work with people on Twitter that they perceive to be more influential than others. From brands who want to know how influential somebody is through to people wanting to work out who to follow and why. Getting real data and statistics direct from Twitter would be very useful - real data on mentions and discussions that go back further than the current search and third-party tools; real data on links and click-thrus and real data on how users use the service.

This is data that Twitter has, it logs everything that everybody does. And data that third-party providers are currently trying to access or just to model and estimate to provide services from Klout (which claims to measure influence) to the search and social media monitoring tools that track mentions and conversations. As Twitter grows and develops they should release more applications and tools that use this - providing us with official views on influence and on the other analysis people are looking for from Twitter. The response from Williams this week suggests that they are using such data internally, and that they could develop external tools to expose this data as a service to others. Although he admits that any such tools would “need to evolve quite a bit more”.

There is a real need for more data from Twitter. As a tool it is changing the way we interact with content, and with others, and often existing analytics and measurement tools just don’t do the job. Take a simple measure such as page views or clicks through from links in Tweets. Many of us grappling with data such as this are uncertain as to whether links, or for that matter photos, that are viewed in a Twitter App (such as Twitter for iPhone), or indeed on the new Twitter website, are recorded in a consistent way along side views of that link through a browser on the original site. The data is not clear and the discussions are confusing.

Twitter has a lot of data as every action we do is recorded. Using data like this can be comlicated but the signs are that they are developing tools that help them internally. It would be great to see these developed and then used externally so we can all be confident that we are getting the most accurate, and the most insightful, analytics we can.