CorpComms Magazine

Would certain Arab governments still be in situ if local activists had not used Twitter to mobilise protestors and raise awareness of their cause internationally? And would certain errant footballers have managed to preserve their anonymity through super-injunctions in the absence of Twitter?

 

In just five years Twitter has attained than 100 million active users, roughly equivalent to the population of the Philippines, who send more 250 million tweets daily.

 

The dramatic growth rate, and the apparent power of Twitter in generating a buzz around issues, both frivolous and serious, has meant that many companies now prioritise influential accounts on social media. But critics are increasingly concerned that some companies are pandering to stakeholders on social media at the expense of other more influential stakeholders who exert influence offline.

 

They argue that some Twitter debates are often no more relevant than drunken debates in pubs, and that some companies assign too much importance to the supposed influence of these social media ramblings.

 

Who is influential?

 

But how do companies determine which tweeters are influential? Certainly, there are plenty of platforms that claim to be able to identify influence, and many, such as Klout, Kred and PeerIndex, seek to measure social media influence.

 

These have their origins in network science, which is the study of how things interconnect, and as a discipline has long been used by everyone from governments - to analyse terrorist networks - to financial services firms looking for perpetrators of insurance fraud.

 

Analysing network structures helps in the search for relationships between agents. In the context of online content, it has the potential to identify where influence takes place by looking at how conversations spread and where they originate.

 

These platforms will often look at indications of influence, such as retweets or the number of followers, but critics claim that considering influence in this way has its pitfalls.

 

For example, Klout generates absolute scores of influence but takes little account of the specific area in which the individual operates. This creates strange anomalies. For example, historically the President of the United States, Barack Obama, has the same Klout score as pop icon Lady Gaga, indicating they are similarly influential. Critics call these scores unhelpful as influence tends to be relative to a particular niche.

 

Rob Blackie, UK managing director of Blue State Digital, says: 'Industries are so niche and complex that this is not a meaningful metric.' Even Klout's co-founder Joe Fernandez agrees. 'When you think about it, the idea of measuring influence is kind of crazy. Influence has always been something that we each see through our own lens,' he says.

 

Other critics claim these tools also suffer from what they refer to as the Warren Buffett problem. Some algorithms overstate the importance of social media activity when there are plenty of influential people who do not have a Twitter account. 'A senior civil servant may be hugely influential but there may be no indications of that online,' argues Blackie.

 

But, despite these criticisms, these tools are increasingly being used for a variety of reasons. In the US, for example, Klout scores are often used as a way of vetting prospective job candidates.

 

Algorithm approach

 

Proponents claim automated platforms could help to identify stakeholders that were previously unknown to companies, but others are sceptical. 'It is true the web has lowered the barriers to entry in terms of becoming well known in a niche area,' says Blackie, citing blogs like that of Guido Fawkes. 'But the stakeholder structures have largely remained the same.'

 

Beyond the free services, there are myriad paid for products, using proprietary algorithms, offering ways to slice and dice the data. Some look at social media alone while others use a combination of traditional and social. They also produce different outputs, from a list or a searchable database to a diagrammatic representation of the influencers.

 

So what are the opportunities?

 

Dan Purvis, director of PR at global monitoring firm Meltwater, thinks these paid for products can help companies gain competitive advantage. 'It's helpful when seeking to curate an online community of brand ambassadors or simply for helping create a more positive impression,' he says.

 

Some tools say they can also measure sentiment continually and see how different stakeholders respond to corporate messages on a real time basis. But it doesn't have to be on social media, argues Flemming Madsen, founder of Onalytica. His service offers a view on how influence passes between think tanks, social and mainstream media, blogs and forums.

 

Madsen suggests that clients think about influence in different ways. He likes to look for stakeholders with 'over influence'. These are the stakeholders who influence the 'influential' stakeholders. In other words, if The Guardian is seen as influential on health issues then who is the newspaper citing in its articles?

 

The source of influence

 

Madsen explains: 'Companies can leapfrog the stakeholder that is well-known and influential by going straight to the source that the reporter is citing. This is how we describe someone who is influential without being well-known. You want to find these people as they will have less demands on their time.

 

'The other metric we look at is centrality: to what extent are you an exclusive source for a constituency of your own? If the BBC, Daily Mail and The Guardian all cite the National Obesity Forum then you have centrality.'

 

The kind of solution and approach will depend on a company's sector and its specific objective. In some, it is easy to identify patterns of influence and interconnectivity: magazines such as Top Gear and presenters such as Jeremy Clarkson hold overwhelming sway in the motor industry, for example. In other sectors, and on other issues, patterns of influence can be more difficult to discern. 'Clarkson has influence in the simplified view of the world. There are examples out there like him, but he's the exception rather than the rule,' says Philip Sheldrake, founding partner at marketing and PR agency Meanwhile, and author of The Business of Influence. 'Every other industry deals with complexity and chaos and yet in this one there is often a tendency to simplify everything and I think it's a mistake. You can't just take the top ten influencers on Twitter and say Job done.'

 

A combined approach

 

Madsen thinks the data is now there to better test hypotheses and produce more evidence-driven communications campaigns. 'Things are more measurable than you might have previously thought,' he says. The question is how to use the bewildering array of tools and data to best effect. 'I would always start with the objectives of the client rather than using the industry or an issue as a starting point. I'd then test out five of those platforms and figure out which works the best,' Blackie advises.

 

Despite all the help available, the difficulty of coming up with a helpful definition of influence remains. 'Influential people must change the beliefs and actions of others,' says Sheldrake. 'And whatever people say we have no digital way for knowing when that takes place.'

 

He believes all the noise in the social media space is a major stumbling block for companies. 'Organisations are at different levels of understanding. There was a peak of excitement but we are now heading towards a trough of disillusionment as it's hard to demonstrate effectiveness of online engagements - especially in this weird economic environment we are currently experiencing,' concludes Sheldrake.

 

Blackie is also sceptical about some of the products being marketed to companies. 'There are lot of people selling rubbish at the moment and this is symptomatic of a broader problem with digital,' he says.

 

Constant refinement