In a world where a tweet or a website posting can torpedo a brand, kibosh a deal or bury a career, the race is on to find a common currency for trust
By Colin Brown
You don’t need to be BP, Toyota, News Corp or even Tiger Woods to know how quickly ‘brand equity’ can be destroyed by screw-ups, cover-ups and indiscretions. In the face of such PR fiascos, the accepted repair strategy has been to come clean as quickly and as contritely as possible. But in an era when corporate skeletons are subject to constant social media scrutiny, your business can suffer just as easily by being too transparent. Ask Bank of America.
Faced with new US regulations that limited the fees that banks can levy merchants every time a consumer makes a debit card purchase, the nation’s second-largest bank could have done what so many corporations have done in the past and made up the loss in hidden charges. Instead, it announced in October that it was going to charge customers a $5 fee for using their swipecards – prompting a huge customer outcry. When rival banks backed away from extorting similar surcharges, Bank of America suddenly caved in and dropped the controversial fee. Any revenue the bank was hoping to raise by the hike was simply not worth the damage to its reputation.
But it’s not just company integrity that is on the line. With Twitter, Yelp and Facebook fast becoming the printing presses of our age, people’s individual brands are being subjected to the same harsh spotlight usually directed at celebrities. There are no safe havens anymore, not even in the workplace, where we have historically hidden behind our professional masks.
Paradoxically, this exposure also opens the door to new business opportunities since there are fortunes to be won and lost on measuring – and safeguarding – one’s standing. Particularly when more and more of our interactions take place online between total strangers.
“I believe reputation capital will become a cornerstone of the 21st-century economy, more powerful than our credit histories,” says Sydney-based author and collaboration consultant Rachel Botsman.
Her book What’s Mine Is Yours: The Rise of Collaborative Consumption provided a vivid demonstration of why trust is vital in a world where social lending, crowd financing, car-sharing and couch-surfing are fast becoming accepted norms. Every day billions of dollars ride on the decisions we make about firms and people, whether it is job recruitment, marketing campaigns, flat-renting, swap exchanges and so on. In making those judgement calls, we place great faith in our own intuitions and those of our immediate social circles. We rely, in other words, on a random accumulation of localised knowledge about people, their backgrounds and various behavioural signals. What if we could pool all those circles of wisdom together and extract a common currency for evaluating everyone’s levels of expertise, social resonance and, above all, such critical attributes as trustworthiness? Well, that race is now on.
“A huge amount of value is created – or lost – by the ability to judge a person’s character and capabilities. Historically most of that data has been siloed in our heads but now a number of tech start-ups are working to bring a ‘reputation graph’ to life that is a partial record of what you know about the people around you and their professional qualities,” says Jon Bischke, co-founder of one of those start-ups, RG Labs. Currently in stealth mode, Bischke’s aim is to build products that change how people hire, form teams and start companies. His firm joins a growing list of start-ups that are using computers to synthesise the trust mechanics that are so second nature to humans in our real-world interactions with one another.
Many see measurement of reputation – trust quotients, if you like – as the next big frontier on the web. Just as Google unleashed the search potential of the internet with its PageRank analysis that assigned a numerical weighting to every nugget of information, so a new breed of reputation brokers is starting to define web 3.0 with the equivalent of ‘PeopleRank’ scores. You might think of these as Yelp ratings for people, creating a hierarchy of individuals and companies based on reputation scores.
Botsman is the first to admit that far more research is needed to ensure that recommendation engines can keep motivating good behaviours while at the same time quickly weed out the bad. There is always the danger, as we have seen with search engine optimisation, that any ranking tool can be gamed or bought. “I do believe the way we form trust face to face can be mimicked online but it’s far more complex then a star, badge or points system. The way trust is measured is too often binary: it’s good or it’s bad. But the fact is, trustworthiness is much messier than that. It will take time, though. And I am not talking about your Twitter ranking – it’s much, much bigger than that.”
Scientifically speaking, however, ‘Twitter rankings’ is pretty much where we are right now. New companies with names such as Klout, Kred, PeerIndex, Identified, PROskore and Twitalyzer, have all developed ways for measuring our social and professional creditworthiness based on the data trails we leave behind us online. They look at our tweets, blogs, web mentions, network reach, online impressions and other social indicators and come up with a rating. These are generally expressed as single scores that fall somewhere between one and 100. Recognising that trust has to be measured across all its different dimensions – and in ways that cannot be easily gamed or bought – such scores are increasingly broken down further to reflect different spheres of influence or expertise.
Exactly which data feeds are harvested – and which computational recipes are used to blend them together – is a hot-button issue right now. As with all gamechangers during their formative years, social metrics have already incited a media and blogging backlash. Scores have instantly been dismissed as vanity metrics that simply play to our egos and competitive instincts; but at worst, they are seen as blunt instruments that can cost people future jobs or else cause irreparable damage to one’s name and communal standing (see Identity Crisis, below). In both cases, the overriding concern has been about lack of transparency when it comes to scoring.
“It is so important not to black box this,” says Damian Kimmelman, a New Yorker who interned on Wall Street as a risk-assessment analyst before moving to London and starting Duedil. His free service offers inside information on every company in the UK and Ireland, and their directors, based on a combination of public records and 900 news websites. Ten billion pieces of information on everything from litigations to financial results and credit checks are indexed and presented in an easily digestible format that comes complete with ‘red flags’ – the first step on the long road towards a reputation system for businesses. “I think I have a contrarian view on all this: no one metric is useful. Any single score, in and of itself, can be very deceiving. Just with statistics, you need to take scores in context. Information can always be misinterpreted. Any reputation system needs to be a two-way street – there has to be an element of trust. But trust itself is often misunderstood. You can have trust between two paedophiles or between two Mafiosi. People often confuse trust and trustworthiness.”
An early target of criticism has been Klout, whose scores have been used since 2009 by corporate brands and their marketers to pinpoint ‘influencers’ worthy of being showered with invites and other freebies – a virtual velvet rope for the twitterati set. This autumn it was Klout’s own social reputation that suffered a double whammy, a victim perhaps of tall poppy syndrome. First, a sudden change in algorithms caused a near-riot among those users who suffered dramatic drops in their Klout scores without any detailed explanations about the revised methodology. Then, following a New York Times article in November, Klout had to apologise for unwittingly creating profiles of minors. Klout, which relies on data from Twitter, Facebook, Google+, LinkedIn, Klout’s founder and CEO, Joe Fernandez Foursquare, YouTube, Instagram, Tumblr, Flickr and Last.fm, as well as blog activity, promised to work with those platforms in introducing safeguards to protect against such intrusions.
“I get why Klout can rub people the wrong way. We are putting scores next to people and that can be initially off-putting,” responded Klout’s founding CEO, Joe Fernandez, in a blog designed to put out the firestorm. “We are not elitist jerks but just a bunch of data nerds passionate about understanding the impact of every person online. We believe that every person who creates content online has influence on some topic, to some group of people.” In Fernandez’s San Francisco worldview, Klout is a “great equaliser”, a way for marketers to recognise and reward people based on the power of their social media voices, rather than just the size of their wallets. He claims that more than 250,000 Klout Perks have been delivered so far.
Nonetheless, rivals have sensed a market opportunity. PeopleBrowsr, a social analytics company that measures trends and customer sentiments for marketers, has just created its own social scoring system known as Kred to measure influence in online communities connected by interest. The emphasis here is on full accountability: users are told exactly how their scores are calculated, even to the point of drilling down to every ‘retweet’ to see how many points it was worth.
“The best way to fight misinformation is a policy of transparency. It’s tremendously important that everyone can understand what their score means, which is why we let everyone see all the actions that go into calculating their Kred,” says Jodee Rich, the Australian digital anthropologist and tech entrepreneur who founded PeopleBrowsr in 2007.
“Kred uses social data to simulate the foundations of strong real-world relationships: trust and generosity within small, close networks of friends and subject-matter experts.
We use indicators like who a person interacts with and whether their content generates action to determine a person’s influence,” explains Rich. And, like so many others in the field, he believes this is just the beginning. “Genetics started with Gregor Mendel experimenting with peas in his garden, and today influence measurement is at a Mendel stage where our understanding evolves every day.”
That recruiters, and not just marketers, are using the likes of Klout to benchmark potential job candidates – or even turn them down, apparently – shows just how quickly social scores are also entering the business domain. It’s a shift that another San Francisco outfit, Identified, is directly targeting with the recent launch of what it calls “the world’s first professional stock market”. By tracking and analysing what thousands of companies are searching for online, Identified provides users with a dynamic measure of how companies rate their professional background. Such real-time feedback is delivered in the guise of the Identified Score. But that’s just the beginning. There is no reason why future data analytics couldn’t go on to provide an automated answer to those high-school polls that ask classmates to vote on who’s ‘most likely to succeed’.
Identified’s scoring algorithms, says co-CEO Brendan Wallace, “still really only scratch the surface of what is possible. We quantify a professional’s desirability along one dimension: recruiting. But the possibilities of how far one could take online personalisation are endless. With enough data, could algorithms predict who’s likely to start a company, who’s likely to be the most creative, who may have the leadership qualities to be the company’s next CEO? Obviously one has to be careful about how much we can rely on data to measure people professionally; being too dependent on data can limit the risk-taking, creativity and synergies that can lead to great organisations. On the other hand, predictive data can be very empowering: in a world of unlimited data, tools like Identified can help companies distill a signal from amidst all the noise that is created by products like LinkedIn, resume databases, and job boards.”
Of course, not everyone makes the right noises – not even the greatest authorities among us. Azeem Azhar, the former Reuters innovation chief who founded PeerIndex, dubs this the “Clay Shirky problem”, referring to the influential internet theorist who, ironically, has a terrible social score. “Every two years Clay writes a phenomenal book – but he does not tweet. He is manifestly an outlier,” says Azhar. “Conversely there are those who get others to tweet for them, essentially creating a special purpose vehicle to enhance their ratings. But then this happens with every index. Look at CDOs (collateralised debt obligations): what seemed so fruity and fresh when rated was really just toxic sludge.”
PeerIndex’s own scoring system tries to account for such anomalies by trying to build a composite picture of an individual and their varying attributes. As with Klout, there’s the iconic overall score, but alongside are also individual measures for activity, audience and authority. Then there are ‘topic resonance’ scores that measure a list of discernible interests related to that individual. In Azhar’s own case that would be: peer influence (score of 50), entrepreneurship (41), customer relationship management (15), start-ups (15) and economics (41). The intended result is an ‘authority fingerprint’ on a topic-by-topic level. PeerIndex’s data sources, beyond the customary social media platforms, include Quora, a question-and-answer website and mobile application that uses its own PeopleRank system for prioritising responses and tracking the best contributors.
Finding fault with such scoring systems is easy, especially for a media establishment that has an instinctive aversion to automation. But are their imperfections any worse than our tried-and-tested methods for weighing up our peers? In all likelihood, no. “Historically every grading system has had its problems,” suggests Azhar. “But I believe the creation of a reputation graph will end up making us all more authentic, and the reason I say this has to do with signalling. In the past the only real managerial capabilities we could rely on were education.
An Oxford degree or a Harvard MBA were the gold standards and yet that system is not very granular. In fact, educational backgrounds are very lumpy and rather unrefined. It’s like the dance of the seven veils where we only get to see a bit of leg. The signals that are produced might overwhelm others that are much more authentic.”
Echoing this optimism is Rachel Botsman. “At the end of the day, I do believe technology can enable trust between strangers and that the tools to do so will get more sophisticated.
It’s staggering, the number of transactions that already happen through collaborative consumption marketplaces, and 99% of all trades are positive. Why? Most users I have spoken to talk about ‘transparency’ and ‘accountability’. They say, ‘I trust you, more than a big bank.’ It’s very much tied to industries becoming human again. The amazing thing is that people behave better, they behave differently when they are interacting with marketplaces based on meaningful relationships and not faceless transactions.”
With consumer power in the ascendant, businesses look upon social media with the same ambivalence that celebrities do. The same mobile megaphone that can trumpet their successes across a global audience can suddenly turn hostile. And back again.
Consider Ashton Kutcher, whose eight million Twitter followers suddenly turned on the techie actor after he inadvertently tweeted a late-night message defending a US college football coach, blithely unaware the coach had been fired in the wake of a child sex-abuse scandal. Or Bank of America, which found itself ‘brandjacked’ by hacker pranksters who created a bogus profile on Google+ that took swipes at its government bailout and its failed attempt to charge debit card fees.
Or Salman Rushdie. Having spent years dodging the fatwa limelight by living incognito, the British novelist found himself in the odd position of fighting to be recognised by Facebook by his commonly known name – and not by the first name, Ahmed, that appears in his passport. “Where are you hiding, Mark?” demanded Rushdie in a tweet aimed at Facebook’s chief, Mark Zuckerberg. “Come out here and give me back my name.” In the end Facebook relented, buckling to pressure from Twitter – the very same microblogging site that Rushdie had also fought earlier in an effort to reclaim his name from an imposter who was using the Twitter handle @SalmanRushdie.
No wonder a survey of 186 executives across Europe revealed that almost half are paranoid about the potential risks to corporate reputation posed by social media platforms. Two trade bodies, the Federation of European Risk Management Associations and the Institute of Risk Management, found that while two-thirds had created formal guidelines covering employee use of social media – and a further 14% were in the process – they felt exposed by web 2.0′s social media tools. Theft of company information and damaging social media buzz were among the big fears.
Joe Public, of course, would argue that it is he, as the ordinary individual, that is being stalked these days by everyone from behaviourally targeted advertisers to insurance companies, law firms and headhunters that are digging up online indiscretions as part of their due diligence.
Research conducted in 2010 found that 70% of US recruiters surveyed had rejected candidates because of information found online such as photos, discussion board conversations and membership of controversial groups. They assumed, of course, that they had targeted the right Joe Public in the first place. Just as companies have to combat negative reviews, so individuals face the prospect of inadvertent discrimination based on false information and even mistaken identity.
“How do we ensure our digital identities reflect our real-world identities?” worries social innovation expert Rachel Botsman. “How do we ensure people can’t steal or intentionally sabotage our reputation capital?”
Among several suggested approaches is one that mirrors the Fair Credit Reporting Act in the US. There, consumer-reporting agencies must provide everyone with one free credit report a year – giving them a chance to dispute any data – and are barred from retaining negative information such as bankruptcies, defaults or tax liens for more than 10 years. In much the same way, people should be able to periodically wipe their social reputations clean.
Among those who would like to see such accountability is Max Drucker, whose start-up Social Intelligence is hired by companies to conduct social media background checks. “The internet, through social media, creates a great opportunity for people to represent themselves in a far broader, more three-dimensional way,” says Drucker. To ensure that, his prepared reports remove references to a person’s religion, race, marital status and disability rather than expose companies to allegations of unfair treatment. Equally, candidates themselves must give their consent and are notified of any adverse information. Above all, Drucker makes absolutely sure that identities are verified – something that cannot not be guaranteed through simple Googling. Rushdie would certainly approve.
The alternative is to take ownership of our online identities through personal lockers that analyse our data for our benefit, rather than be held hostage to the whims of third-party diagnostics.
A New York start-up, Singly, is among those building an open-source platform that would allow users to gather in one place their entire social footprint, from clickstreams, photo tags and check-ins to the heart-rate data monitored by our running apps.
“We want to help people own a copy of all the data they create – social, personal, public and private information about relationships, interests, preferences, locations, purchases, activities and more, and empower people to do more with their information,” suggests co-founder and CEO Jason Cavnar. “The concept of a reputation economy is really a leading indicator for a much bigger shift that is taking place: people are becoming aware of the concept that their data is a currency of sorts – that it is valuable – and it is being used as a proxy for worth.”
It’s high time, in other words, that we start stalking ourselves.