Have you ever written a fake review on Yelp, Amazon or the App store to help promote a friend’s restaurant, new product or app? Did you think that it was’nt a big deal and that your comment wouldn’t hurt anybody? If so, now is the time to reflect and change your habits.

What is at stake:

There are two reasons explaining why fake reviews are dangerous for online businesses, whose models rely on trustworthy assessments of products and/or services:

1. Fake reviews kill transparency: by providing fake comments, one essentially removes any element of honesty and trust, which is at the cornerstone of ecommerce businesses. Take Amazon for instance. A fake product review creates friction in the supposedly seamless transaction process by providing false pieces of information. Amazon ends up with unaware customers, who might just end up purchasing subpar products.

 2. Fake reviews limit network effects: Amazon’s value proposition is offering an independent and neutral third party platform to sellers and buyers willing to do business with one another. If that platform becomes crooked and biased, and if sellers can effectively push their products with paid reviews, one can easily imagine online shoppers leaving the platform for better sources of information. Fewer shoppers, fewer sellers, fewer transactions, and in the end, a weakened market place platform with no network effects.

The Amazon police:

In an effort to cut down the risks associated with such threats, major online businesses that rely on ratings and reviews to operate have pulled out the big guns. Amazon is the best example:

Since April 2015, Amazon has launched a very aggressive campaign against fake review providers. It started by filing suit against the operators of buyazonreviews.com, buyamazonreviews.com, bayreviews.net and buyreviewsnow.com. Before these sites were taken down, they allowed any interested Amazon seller to buy fake 4-5 star customer reviews in order to boost sales. As for the sellers who commissioned fake reviews, Amazon banned them as well.

Yesterday marked the second step in Amazon’s crackdown of fake reviews and in its fight against those who create a poor ecosystem. This time, Amazon went directly at those using Fiverr.com to buy and sell Amazon reviews. In a nutshell, Fiverr is an online marketplace that allows users to offer small tasks and services for USD 5. Services include writing, editing, or programming, among others. With the help of Fiverr, Amazon spotted over 1,100 fraudulent individuals and sued them all with the hope that this will send a strong signal to those who try to play around its terms of service.

What more can be done:

With that said, small businesses do not have Amazon’s strike force and it is often too expensive and inconvenient for them to file suits. How can these smaller players combat fake reviews? I see a few ways smaller players can fight fraud from flourishing:

  • Make the review writing process more demanding by asking a series of personal questions that can help identify writers. This can help remove robots from writing the reviews.
  • On one hand, manually filter reviews and remove fake looking ones. This is time consuming and not perfect though. On the other hand, create algorithms that detect fake reviews.
  • Identify dishonest sellers, writers and tag them publicly as such on their profile. Run a zero tolerance policy for everyone to see. Yelp, through its consumer alerts program, along with TripAdvisor among others, are using this technique for instance.






By: Edouard Delvaux

read more

The Story of TripAdvisor

In an unassuming building in Newton, Massachusetts, the team at TripAdvisor is working hard on perfecting the updates to the code to be released on Monday. “Speed wins,” the sign on CEO Steve Kaufer’s door reads. Even now, the $11 billion company still strives to retain a start-up’s sense of urgency and a culture of modesty that speaks much to its founding.

Establishing TripAdvisor

As recounted by Kaufer, the original business model for TripAdvisor was to be a white label search engine for web portals to the likes of Lycos and AOL. The company initially hired editors to comb through archives of professional travel articles on content provider websites, such as The New York Times, in an effort to sell this database on a per query basis. The value proposition was to provide high quality, curated content that would be, as Kaufer had hoped, costly to replicate.

Yet a year and a half later, TripAdvisor had only closed one licensing deal with Lycos. As it turned out, popular web portals and Kaufer’s team fundamentally disagreed on who should be the one to pay. Like many innovative start-ups with unproven standalone value, TripAdvisor was fighting the age-old problem of, “We seem to be doing OK without [your product]. Why do I have to pay you for it?” Then, unimaginably, 9/11 hit and the entire travel industry grinded to a halt.

In the midst of despair, Kaufer noticed that the demo site TripAdvisor.com was picking up traffic. Desperate to monetize, the team experimented with a number of advertising methods and finally settled on a cost-per-click model similar to Google’s search ads. In a way, TripAdvisor benefited from its position at the nexus of an extremely valuable user base – those with high purchase intent for relatively high priced services – and the hospitality industry’s low marginal cost economics. Then, in a serendipitous move, TripAdvisor added a button that allowed users to provide their own reviews. To the founding team’s surprise, user-generated content expanded exponentially, quickly replacing the edited content on which TripAdvisor had sought to establish itself.

Scaling Traffic via SEO

With the rapid growth of user reviews, a number of complementary effects took over. Unsurprisingly, TripAdvisor relies heavily on Google to bring in traffic, and thanks to its remarkable SEO [search engine optimization] strategy, TripAdvisor manages to show up – front and center – on just about every travel-related search. This is no coincidence. The team at TripAdvisor has made a number of highly effective choices that have helped the destination site retain its top position in the industry.

First, TripAdvisor’s dedication to SEO has created a virtuous cycle in which its advantageous search rank helps the company to achieve higher share-of-mind for potential and repeat users. Then, as new users are welcomed into TripAdvisor’s vibrant community, they often feel obligated to give back to the community that has helped them plan such a wonderful trip in the first place. With more high quality and fresh content that appeals to the algorithms of major search engines, TripAdvisor will continue to cement its placement at the top of travel searches, thereby capturing more content-generating traffic.

In addition, TripAdvisor offers a number of tools for hotels to promote reviews directly on their websites, with each of those widgets and badges essentially serving as an inbound link to TripAdvisor, consequently boosting the page’s ranking on a search. In return, hotels, especially those reviewed favorably by TripAdvisor users, receive some of the most credible advertising money can buy. As the community expands and brand recognition for TripAdvisor overshadows that for individual hotels, hoteliers find that a TripAdvisor stamp of approval becomes almost a prerequisite for many travelers.

Ultimately, what’s particularly impressive is that despite characteristics of the fragmented online travel industry that are averse to winner-take-all dynamics, TripAdvisor has skewed many aspects of the industry to its favor. By snatching up competitor after competitor and absorbing their users, by creating a vibrant community as well as a truly differentiated product that’s difficult to imitate, TripAdvisor has closed the door behind itself to future competitors. In the end, TripAdvisor’s story is one of rapid pivots and of virtuous cycles, of luck and of thoughtful planning, an inspiration to entrepreneurs everywhere.


“Fact Sheet.” TripAdvisor. Last updated July 2013. <http://www.tripadvisor.com/PressCenter-c4-Fact_Sheet.html>.

“TripAdvisor. Recent milestones.” CrunchBase. Last edited October 24, 2013. <http://www.crunchbase.com/company/tripadvisor>.

“TripAdvisor’s Stephen Kaufer: Founder Stories.” TechCrunch interview. December 21, 2011. <http://www.youtube.com/watch?v=5JBa9ZH1i1M>.

Alex Cocotas. “How $3 Billion TripAdvisor’s Business Works.” Business Insider. December 23, 2011. <http://www.businessinsider.com/how-tripadvisors-business-works-2011-12>.

Andy Gelfond. “TripAdvisor Architecture.” High Scalability. June 27, 2011. <http://highscalability.com/blog/2011/6/27/tripadvisor-architecture-40m-visitors-200m-dynamic-page-view.html>.

Ava Seave. “How TripAdvisor Grows Scale and Network Effects: Expertise in Gathering UGC.” Forbes. June 20, 2013. <http://www.forbes.com/sites/avaseave/2013/06/20/how-tripadvisor-grows-scale-and-network-effects-expertise-in-gathering-ugc/>.

Brian Fitzgerald. “How The Travel Industry Is Working Together To Help TripAdvisor’s SEO.” Orourke. May 15, 2012. <http://www.orourkehospitality.com/blog/how-the-travel-industry-is-working-together-to-help-tripadvisors-seo>.

Christopher Williams. “Interview: TripAdvisor founder Steve Kaufer.” The Telegraph. April 2, 2013. <http://www.telegraph.co.uk/technology/social-media/9960224/Interview-TripAdvisor-founder-Steve-Kaufer.html>.

Jeffrey Bussgang. “The Secrets to TripAdvisor’s Impressive Scale.” HBR Blog Network. October 2, 2012. <http://blogs.hbr.org/2012/10/the-secrets-to-tripadvisors-im>.

Jessica Livingston. “Founders at Work: Stories of Startups Early Days.” APress. September 19, 2008.

Rama Chanani & Jackie Kelly. “An Interview with Stephen Kaufer, Co-Founder & CEO of TripAdvisor.” Harbus. April 1, 2013. < http://www.harbus.org/2013/an-interview-with-stephen-kaufer-co-founder-ceo-of-tripadvisor/>.

Robert Weisman. “TripAdvisor shows small firms the way to success.” The Boston Globe. January 17, 2005. <http://www.boston.com/business/globe/articles/2005/01/17/tripadvisor_shows_small_firms_the_way_to_success/>.

Roderick Ioerger. “TripAdvisor Growing With Help of User Generated Content.” Marketing Pilgrim. May 24, 2007. < http://www.marketingpilgrim.com/2007/05/tripadvisor-growing-with-help-of-user-generated-content.html>.

read more

Online reviews are a great tool to facilitate our decision making and have massively gained in popularity over the last years. Getting feedback about the cleanness of a hotel room, the quality of service in a restaurant or user experiences with a new laptop has never been so easy. Reviews reduce our search cost, convey information that we would otherwise not get or highlight issues, we did not think about reading the glorified manufacturer’s product description. I recently bought a mixer online and based my decision on product reviews on Amazon, videos on youtube and online comparison sites. In class we discussed, how reviews are the most powerful way to create trust and reduce risk for SaferTaxi and airbnb. Many positive comments on a taxi driver or potential tenant indicate that this person might be trustworthy. Yelp and other review sites have built massive businesses around providing these services.

However, businesses have also understood how valuable or damaging reviews can be to their success and started to actively game the system. Here are three reasons why you should be extremely careful using online reviews.

Up to 30% of all online reviews are fake! (Weise, 2011) Research has shown that for a broad variety of products, a large share of reviews are not posted by genuine users, who have actually used the product or service and want to share their honest feedback. For example some company employees and paid “freelancers” write about amazing experiences with the intent of misleading potential customers. Because of the scale of this issue, the US Federal Trade Commission has passed a guideline that requires online reviewers to disclose any affiliation with the respective manufacturer or service provider. Along these lines the UK government has ruled that TripAdvisor is no longer allowed to advertise “honest, real or trusted” reviews from “real travelers”, because the company cannot guarantee that the posts are not fraudulent (Mayzlin, Dover, & Chevalier, 2012). However, these guidelines are hard to enforce. Companies like Microsoft, Google, Yelp and TripAdvisor are very concerned about deceptive reviews and have sponsored research on computer algorithms to detect and remove such posts.

Reviews are systematically skewed! Why does Yelp actively discourage businesses from asking for reviews? They do it because there is a strong selection bias (Yelp, 2010). Companies have an incentive to only ask satisfied customers for feedback and sometimes even give away merchandise for 5-star ratings. This generally inflates reviews and skews the picture (Streitfeld, 2012). Small hotels are particularly active in review fraud. Research shows that they are 10% more likely to have top ratings on TripAdvisor –open to every user- than on Expedia, where the reviewer had to actually book and pay for a stay before writing a review. It also points out that large hotels are systematically more likely to receive some very bad feedback if they are located near a small independent hotel. The authors attribute this to active discretization and show that the effect is stronger on more liberal review sites like Tripadvisor (Mayzlin, Dover, & Chevalier, 2012).

We are bad at detecting fraudulent reviews! Although we think we can spot the deceptive reviews, the chances of getting it right and not being fooled are very low. Even if we know that one of two postings is fake, on average we only get it right 60% of the time, which is hardly much better than flipping a coin (Ott, Choi, Cardie, & Hancock, 2012). Automated algorithms perform with up to 90% accuracy. For your next hotel stay, try out www.reviewskeptic.com. This tool helps you identify fraudulent reviews based on research at Cornell University.

Reviews are still a great way of making better purchase decisions with limited time and exposure to the actual product or service. However, we should not blindly trust them and know about the pitfalls. As a general rule of thumb,

(1) reviews from sites that require proof of past usage,

(2) posts from users, who have written about multiple brands over a longer time and

(3) testimonials without extreme notions (e.g. excessive use of superlatives or opinions diverging vastly from others) without certain keywords (e.g. husband)

are more likely to be legitimate ones. For more detailed tips I recommend the Cornell research.


Mayzlin, D., Dover, Y., & Chevalier, J. (13. August 2012). Promotional Reviews: An Empirical Investigation of Online Review Manipulation. Retreived from: Social Science Research Network: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2128860

Ott, M., Choi, Y., Cardie, C., & Hancock, J. (2012). Finding Deceptive Opinion Spamby Any Stretch of the Imagination. Ithaca, NY: Cornell University.

Streitfeld, D. (26. January 2012). For $2 a Star, an Online Retailer Gets 5-Star Product Reviews. Retrieved from: The New York Times: http://www.nytimes.com/2012/01/27/technology/for-2-a-star-a-retailer-gets-5-star-reviews.html?_r=2&hp&

Weise, K. (29. September 2011). A Lie Detector Test for Online Reviewers. Retrieved from: BloombergBusinessweek: http://www.businessweek.com/magazine/a-lie-detector-test-for-online-reviewers-09292011.html

Yelp. (13. August 2010). Yelp WEB LOG. Retreived from: Don’t ask for reviews:



read more