Since the dawn of the web, community boards, forums and review sites have been affected by the problem of bogus reviews. While users increasingly expect to see reviews of businesses before making purchases or hiring decisions, they are becoming ever more wary of the opinions expressed online.
A recent Wall Street Journal headline blared “Fake Reviews Are Everywhere,” citing eye-opening research from professor Michael Luca at Harvard Business School that suggests at least 20% of reviews on Yelp are bogus — and possibly more. Even Amazon, often considered the gold standard of review quality is not immune to the problem. A 2011 study by a team at Cornell found that of Amazon’s top 1000 reviewers — as rated by other users — fully 85% (!) received free products but did not disclose this potential source of bias.
Given the scope of this problem, how do we in the business of local search begin to change the tide of trust? How can we help the end consumer while still improving the local search experience? To answer these questions, I will first unpack the problem, then present a framework for where trusted reviews matter most, and finally provide some suggestions for what can be done by those in the industry to address the issue.
First, let’s examine the often unarticulated concept of what makes reviews so powerful. It is axiomatic to say that reviews are helpful when searching for local businesses, but in fact, their most important function is driving the user to a purchase decision — it’s the choosing that actually matters. In an age where data is everywhere, local listings are table stakes – anyone can find a provider. What matters today is giving the user an effective tool to narrow down large lists of potential businesses, to the one or several candidates that are the best fit.
And when seeking fit, not all searches are equal. In some cases, finding a local business is primarily an exercise in matching location and capabilities — for example locating a nearby gas station or ATM. Reviews start becoming important when users require advice or guidance. When searching for a restaurant or gym, honest opinions can prevent an inconvenient, or unpleasant experience. Even more important, when seeking a mechanic, or pediatrician, the downside of a bad choice can be severe.
As a consequence, reviews for just these kind of businesses have exploded, and the trend has not been lost on SMBs. More than ever they understand the impact of reviews on their business (“A half star increase in Yelp ratings translates into 19% increase in sales”). So the incentives are pushing those with a vested interest to game the system, and this vicious cycle is beginning to erode people’s trust. Indeed a Maritz study from just last month suggests that among users of review sites, between 40%-60% of people do not trust the opinions they read! (“Confidence Crisis: 40 Percent Don’t Trust Online Reviews”). That’s 40% of the users of review sites – imagine how other people feel!
While this crisis of confidence continues to play out, many have been left to wonder whether there is any solution. Academics have tried developing algorithms to root out fake reviews, and some people claim they can pick out the good from the bad. Neither approach is likely to change the game; they simply don’t work.
There is, however, reason for hope. In fact, there is a solution that has already started to work across a variety of review and comment sites. What is this magic bullet, this panacea? In one word: identity.
Consumers make decisions based on trust, and just as trust is established in the offline world by people that are familiar or known to each other, so too must the principle of trust be extended online. That means reviews, opinions and recommendations must be submitted with identity. It is easy to spot the trend already in progress; a majority of news sites now require Facebook login to comment, and Google’s recent announcement of including identity in ads is one more clear indication that trusted opinions sell.
It’s true that pundits have been touting social as the savior of local search for many years. In fact, it reminds me of the voice recognition world, back when I worked in that industry in 1999. People loved to say at that time that voice recognition had been “a year away” for the past 20 years — now it’s been nearly 15 years since then, and Siri and her siblings are still working to gain credibility. So what’s different now for social media and search? Is there any way we can actually achieve social, local search, and help consumers make informed decisions about local service providers, and on the way, unlock the massive potential in this market?
The answer is a resounding yes. If you haven’t already, provide a way for users to establish identity, and with it, credibility. There are many tools to accomplish this, from Facebook and LinkedIn to Google+ and more. Also, try to authenticate the connection between a reviewer and a transaction whenever possible (the New York Times recently profiled a new initiative between TripAdvisor and American Express to do just that – kudos!). And finally, push as deep as possible to drive to the source of authentic conversations. At whodoyou, we try to make it easy for people searching for local businesses to see advice already shared between friends — and link back to the original conversation. There are many examples of companies that are either hosting this dialogue, or building communities where real people can share honest information.
As long as there are economic incentives to create bogus reviews, unscrupulous people and businesses will continue to exploit the platforms that make it easy to deceive. For the local players seeking to help users in selecting a great business, use the offline world as a guide. Require an identity, build communities of real people sharing advice, and give legitimate customers a megaphone to honestly rate the service received. If these steps are taken, we can begin to welcome in a new era of trusted reviews, and perhaps finally reach the holy grail of local and social.
Yoav Schwartz is founder and CEO of whodoyou.