Sunday, 28 April 2013

Bad reputation: blackmail, corruption plague online reviews

Once lauded as democratic, user-generated content is now being exposed as a way to damage hard-won credentials, writes Malcolm Knox.

Illustration by istockphoto.
Lucy Reynolds, an Australian expat with a successful business in Italy, woke up one day to find herself under attack. The Sydney-born former lawyer had spent most of her 30s building up her enterprise, taking tourists on walking tours of the sights of Florence - the Uffizi, the Duomo, the Baptistery of John the Baptist, the Ponte Vecchio - as well as private art galleries, and using her decade's experience of being married to a local to steer them towards her favourite restaurants and cafes. She was listed on TripAdvisor, the world's biggest online travel forum, and was pleased to see on her regular checks of the website that her customers gave her positive reviews.

Overnight, her relationship with the site turned into a nightmare. When she logged onto her computer in her Florence apartment one morning in late 2011, the normally favourable TripAdvisor reviews of her business had been pushed down the page by a screed of short, hostile warnings. She was suddenly untrustworthy and overpriced and the formerly positive word of mouth about her tours had turned poisonous. She began to receive threats from people purporting to be customers. "They said if they did not receive discounts or upgrades, they would damage us on TripAdvisor," says Reynolds. "It was blackmail."
Reynolds's investigations into TripAdvisor, a site that boasts 50 million traveller reviews, which at the time billed itself as "The world's most trusted travel source", uncovered a trail of corruption. The negative twist in the reviews of her business coincided with TripAdvisor having appointed a new DE, or Destination Expert. The site's DEs are the equivalent of site mediators, whose experience and knowledge are meant to act as a guide and filter for customers.
The new DE, who "consistently gave misinformation and denigrated my company", says Reynolds, turned out to be "a disgruntled ex-employee who now worked for our main competitor. This was a pretty terrible situation for us, having a person who wields quite a bit of marketing power on the world's most trusted and visited travel-advice website denigrating my company. What's more, this expert consistently recommended the company she works for. All this was done by not being transparent about who she [was]."After her direct complaints were ignored, Reynolds found a well-populated discussion thread on a private TripAdvisor forum called "How do I prevent extortion?" Many travel companies were making the same complaint as Reynolds - not just that TripAdvisor reviews had turned against them, but that they were being shaken down by customers threatening to blackmail them on the site if they didn't receive favours.
The "How do I prevent extortion?" thread was soon shut down by the site, but Reynolds began legal proceedings against TripAdvisor. Again, she found she was not alone. The high-profile Scottish hotelier Duncan Bannatyne, whose hotel was likened on TripAdvisor to the Fawlty Towers hotel, took TripAdvisor to Britain's Advertising Standards Authority. Its investigation elicited TripAdvisor's admission that it did not check the veracity of its reviews or the identity of its reviewers.
Meanwhile, numerous class actions for extortion were initiated against San Francisco-based crowdsourcing review site Yelp. No such action has succeeded in court, for lack of evidence, though a new action has been started by a veterinary clinic in Long Beach, California, which claims Yelp has asked for $300 a month in exchange for removal of negative reviews. Since its founding in 2004, Yelp has regularly rejected accusations of extortion from businesses who decline to buy advertising space.
Tripadvisor is one of the largest of a flotilla of consumer-review sites based on crowdsourcing, or user-generated content (UGC). Even though many are less than a decade old, they now permeate all corners of the consumer experience. Nielsen Research has estimated that 71 per cent of Australians base their purchasing decisions on UGC reviews that appear on sites such as TripAdvisor and Yelp (local businesses), Amazon (books and other consumer goods) and Urbanspoon (restaurants and cafes). UGC has eclipsed traditional "top-down" sources of authority such as mass-media reviews and mainstream advertising to the extent where, for example, a web-based "people's network" of accommodation, the five-year-old Airbnb website, now fills more rooms (250,000 listings in 192 countries) than the 94-year-old Hilton Hotels chain. Airbnb has no bricks and mortar, but its chief asset, what "social innovator" Rachel Botsman terms "reputation capital", underpins a bigger clientele than Hilton's 540 hotels and resorts.
What is reputation anyway? The writer Frank Moorhouse once answered with a question of his own: "What is truth?" Reputation is fundamental to individuals' identities. As Cassio lamented in Shakespeare's Othello, when he lost his reputation he lost "the immortal part of myself, and what remains is bestial". And yet reputation is also highly contestable. It is neither fixed for life nor free from doubt. Iago replies to Cassio that he has only lost "an idle and most false imposition; oft got without merit, and lost without deserving".
Being both highly valued and contestable, reputation has always been a battleground for corporations. In the 20th century, Coca-Cola's creation of a reputation for its drink virtually gave birth to the modern advertising industry. Organs of mass media, such as newspapers and radio and television stations, "made" reputations - of individuals and institutions - through the reach and authority of their own names. For a moment in time, somewhere mid-20th century, reputation seemed to be something that could be handed down from the mount of the ABC, The AgeThe Sydney Morning Herald or the Commonwealth. Yet where power was centralised, opportunities for manipulation arose. But no sooner was the process of reputation-making "settled" than the counter-process of reputation-faking came into being, through the darker arts of public relations and image cultivation. The history of the century was a story of rising scepticism, and by the time the internet arrived, whole populations were primed to disbelieve.
In a blink, social media inverted old ideas of reputation. Crowdsourcing held out the seductive idea that we, the people, now control the truth about whether companies and their products are good or bad. Advertising and reviews in mass media are no longer the final word; we seek information from each other. Consequently, businesses flocked into social media to preserve and promote their "reputation capital". But if anyone imagines that UGC has created a level playing field of honest, uncorrupted feedback, they should hear the warnings of those, such as Lucy Reynolds, whose reputations have been torn apart. While Reynolds endorses the concept of sites that "tip the balance of freedom of consumer information" away from proprietors and towards customers, she was so scarred, emotionally and financially, by her experience on TripAdvisor and she lives in such fear of reprisals that she asks that her real name not be published.
"TripAdvisor takes reputations as a kind of cyberspace real estate and squats on them, captures them and then profits from them," she says. "Business owners are held to ransom by unscrupulous persons interacting anonymously."
In fact, social media have made the question of reputation more susceptible to corruption rather than less; and it can be argued that the internet, far from democratising the evaluation of "reputation capital", has concentrated it more than ever in the hands of a few giant corporations.
Of all the online reputation-makers, none is as powerful as Google's PageRank, the algorithm that powers its search engine.
PageRank, the first version of which was devised by Larry Page and Sergey Brin in 1996, is devoted to converting word of mouth into a mathematical formula. A simplified description of PageRank is that it defines reputation both by the number of times a web page is visited and the reputation of other pages that link to it. Web pages are assumed to earn reputation in direct proportion to their usefulness and popularity, as measured by the usefulness and popularity of those other pages. No reputation-maker in human history, not even the Catholic Church or the Roman Empire at their heights, has had the reach and influence of PageRank.
For the 17 years PageRank has existed, it's been under assault. The algorithm is as closely guarded as the formula for Coca-Cola or the Colonel's 11 secret herbs and spices and is constantly altered, for the dual purposes of improving reliability and accuracy, and to evade the clutches of the mini industry of search engine optimisation (SEO).
SEO firms work to decode search engines such as Google, and sell this service. A primitive form of SEO was to manually or automatically hit a company's website so often that it would eventually bob up on the front page of a Google search; it was the tech world's equivalent of a record company buying 10,000 copies of its latest release so it would hit number one on the charts. The resultant publicity would, presumably, keep it there.
SEO has evolved into two broad streams - "white hat" and "black hat". White-hatting seeks to flood a website with values that PageRank will reward, whereas black-hatting seeks to trick PageRank by concealing such values within a page. (In the internet's early days, a black-hatting technique might have embedded a highly searched word, such as "sex", on a page invisibly to the human eye but visible to a computer, so that anyone searching for "sex" would come to that page.)
Google has fought back by issuing counter-measures against SEO. The search engine's 2012 "Panda" update tracked and penalised sites that were trying to dupe PageRank. In 2006, Google blackballed the German branches of multinationals Ricoh and BMW for black-hatting PageRank. The war between Google and the SEO industry is ongoing.
Google's claim is to take, through mathematics, the subjective element out of reputation-making. But what if reputation is inherently subjective? How is perception measured?
The reputation index of Australian companies, run since 2000 by research firm AMR, attempts to combine objective and subjective elements. Oliver Freedman, managing director of AMR, explains: "Previous reputation indices [including one run in the early 2000s by Fairfax Media] sought to give a value to intangibles such as trustworthiness and social responsibility. Ours is quantitative, not anecdotal."
The AMR index simply takes the top 60 companies, in terms of revenue, and polls 5000 members of the public on their perceptions of those companies. It does not impose an idea of what reputation should be: the public decides. The difference between popular perceptions now and in the past, says Freedman, is that "it's no longer an assumption that large companies are doing the right thing. They have to prove it."
As AMR's index has become widely publicised - its findings are broadcast and re-broadcast in traditional and social media - it has to respond to companies' efforts to influence it. "It's not a case of companies trying to boost their rating," Freedman says, "so much as trying to take themselves out of it when something bad has happened." Some companies have suffered from unfortunate timing, he admits. "Nobody had heard of Pacific Brands a few years ago, but everyone knew their products [such as Bonds]. The first time the public heard of them was when they started laying people off and announced bad results. So their reputation was built at a moment of negativity."
Before Google, the biggest reputation-makers were mass-media institutions. Although their power's been eroded by social media, mass media are still "incredibly powerful in creating social objects - what people talk about at barbecues", says Iain McDonald, the founder of advertising agency Amnesia Razorfish and chair of the industry's Communications Council Digital Committee. Nevertheless, the interaction and rivalry between mass media and social media is "complicated and fast-changing".
An example of that fractious interaction came last year when a Melbourne software developer, Russell Phillips, began to wonder if online opinion polls, given the imprimatur of venerable newspaper mastheads, were easily corruptible. Phillips had noticed a weakness in the design of polls run by news.com.au and Fairfax Media, and wrote an automated "bash script", a program that repeatedly voted for both sides of online polls. Resultant polls recorded abnormally large voter numbers, always coming out 50:50, as Phillips had programmed. He says it was "a very quick task" to break the polls, but labour-intensive - the province of a specialist - to write the code deadlocking the polls. "It required weeks' worth of late-night coding," says Phillips. "Nobody would code something like that opportunistically - it takes a fair amount of dedication."
Inevitably, some newspapers started to report the polling results as "news". Phillips, who had no intent to sabotage the news, notified his main victim, news.com.au, of what he'd done. He was ignored until he took his story to ABC TV's Media Watch program, which reported the exchange in February.
Phillips's work was inspired, in part, by the international "Anonymous" movement, which, among other campaigns, toys with online trend-making systems to expose their flaws. "A common one is a method to influence the Google search engine by flooding it with searches for a particular term," says Phillips. "The idea is that Google detects the trend and starts bumping that particular search up the auto-complete chain - you know, the one that tries to finish your sentences."
Anonymous and its brother organisation, 4chan, have staged several coups through automated hoaxes, including getting 4chan founder Christopher Poole and North Korean leader Kim Jong-un to the top of Time magazine's "Person of the Year" poll. Anonymous also rigged a poll among schools voting for Taylor Swift to sing for them, so that a school for deaf children won.
Whether the reputation-maker is TripAdvisor, Google or a newspaper masthead, manipulating it is not just a guerrilla activity, it is big business. Social media might seem to be a game-changer in the reputation stakes, but big companies spend serious money trying to infiltrate and distort them. Social media terrify corporations, because they transform over-the-back-fence scuttlebutt into a permanent digital record. James Griffin, partner at social research firm, SR7, explains: "Historically, word of mouth or newspaper articles would dissipate over time. The digital world has given what used to be barbecue talk a more permanent stamp, plus speed and spread. Something that might once have been spoken between a few people can now go worldwide instantly and be very difficult to remove."
To influence this new form of reputation-making, companies quickly established presences in social media. But initially this work was left in the hands, says Griffin, of "the youngest person in the office". An outcome of inexperienced social-media staff led to some notable corporate disasters. ChapStick deleted adverse comments from people offended by its advertisement on Facebook, and received worse publicity for the censorship than for the advertisement itself. Qantas and Target, trying to seek consumer input into future products, instead received torrents of abuse relating to their performance.
Companies learnt that social media was too serious to be left in junior hands. After Toyota had to recall 2.3 million vehicles with faulty accelerator pedals in 2010, its then-president of US motor sales, Jim Lentz, went on the social media network Digg, where he received 3200 questions. The Dell Corporation, in response to blogger Jeff Jarvis's "Dell lies. Dell sucks" public revolt, set up a "social outreach" division in 11 languages, staffed by 70 employees, featuring personal appearances by founder Michael Dell.
But as corporations have seized control of the social-media conversation, regulators have informed them they can't have it both ways. Last August, the Australian Advertising Standards Board (ASB) deemed companies' Facebook pages to be "a marketing communication tool over which the advertiser has a reasonable degree of control". Whatever appears on a company's Facebook site, positive or negative, is now the company's responsibility.
This was overkill, says Iain McDonald. "Digital natives generally know how to separate remarks from promotions," he says. But one knock-on effect of the ASB's ruling is to accelerate a cultural change in how companies manage social media internally.
"The last five years have seen a remarkable shift where social media is now overseen by risk management and audit committees, part of senior managers' responsibilities, not just a young people's thing," says Griffin.
McDonald goes further, saying, "We are moving towards industry accreditation, training and qualification for social-media moderators."
There is also a push in the USA and the UK for governments to work with Facebook and Twitter to create single sign-ons for individuals, verified by birth certificate, passport or driver's licence, to make it harder to create false identities.
"I think in five years' time, when we have legislation, more solid self-governance, privacy laws and automated tools [fixing identity], we'll be looking back on the present period as a time when the whole social media environment was very loose," says McDonald.
The moment in which social media was seen as a truthful, democratic, level playing field for the making of reputations was brief. The University of Illinois concluded from research last year that one-third of all UGC web reviews are now fake, either with the purpose of boosting or denigrating a product. Sites such as Fiverr and Freelance advertise for writers to produce fake reviews. Technology research firm Gartner Inc has predicted that 10 to 15 per cent of web ratings and reviews will be paid for by companies by next year. The practice known as "astroturfing" - companies creating fake grassroots campaigns either supporting their products or denigrating their competitors' - is rife in the travel industry. Last year, wide currency was given to a list of the "world's best airlines" published online by ATRA, the Air Transport Rating Agency. ATRA turned out to be a Swiss start-up associated with one of the listed airlines. Open and democratic social media had a quick moment in the sun before corporations and fakers moved in.
Whether seen as an explosion of consumer freedom and democracy or open slather for reputation saboteurs, social media's "Wild West" era might have passed its peak. But in the new world of reputation-making, old slander can stay visible forever.
Lucy Reynolds never received the satisfaction she was seeking - for TripAdvisor to admit the corruption of its "Destination Experts" system and to delete the defamatory comments. "It is easy to see why many consumers champion the use of sites like TA and the access to consumer commentary they give," she says. "It would seem sites like TA have tipped the balance of freedom of consumer information in many ways to aid consumers in having access to what should be transparent information. But in many cases, the information is not transparent ... The whole point of TA is that it captures, and can control, reputations. That is why it can be so damaging and why it is such a powerful marketing tool. No company listed in it can afford to ignore it."
In the end, unable to ignore it, Reynolds was forced to escape from it, moving her business.
TripAdvisor has now undertaken reforms, including opening a hotline for aggrieved hoteliers, and has changed its slogan. Once the world's "most trusted travel site", it is now the "largest".


Read more: http://www.smh.com.au/digital-life/digital-life-news/bad-reputation-blackmail-corruption-plague-online-reviews-20130422-2i944.html#ixzz2Ro5j1gjJ

No comments:

Post a Comment