Buyer beware: Fake reviews and what you can do

UNIVERSITY PARK, Pa. — Black Friday and Cyber Monday are right around the corner, but before you add a product to your online shopping cart, you may want to check if the product reviews you’re relying on are authentic or fake.

Penn State News spoke with Stefan Wuyts, professor of marketing and director of the Institute for the Study of Business Markets, about the dangers of fake reviews and what consumers and businesses can do to identify and stop fakery.

Q: What dangers do fake reviews pose?

Wuyts: Consumers use reviews for making purchasing decisions. If reviews are fake, then consumers may mistakenly believe that a particular product is very good, whereas the true experiences of users of the product may be more negative.

Take, for example, TripAdvisor. They report that in 2020 they removed more than 2 million fake review submissions from their website because their algorithms flagged these reviews as potentially fake and misleading. They found that in 97% of these cases, the culprit was actually the firm itself. The firm asked their employees or their family members to upload positive reviews or even provided financial compensation to individuals to upload fake reviews.

The Federal Trade Commission (FTC) is coming up with new regulations to try to fight fakery on online review platforms. The FTC made an assessment that if they could manage to fight fakery, it would increase consumer welfare by as much as $15 billion. So from an economic perspective, from a societal perspective and from a marketing perspective, it’s a really big issue.

Also, honest firms are the victims of these bad actors that upload all kinds of positive reviews of their own products, because the honest firms are still part of the competition. They lose sales to these opportunistic competitors. Fighting fakery is in their interest as well.

What can consumers do while online shopping to spot fakery?

Wuyts: First of all, you may want to look at the reviewers. Did they upload few or many reviews at strange moments of the day or at regular times?

Second, you can look at the review itself and treat those reviews that just say “wonderful, fantastic product” with a grain of salt.

Third, I would recommend that consumers look beyond the individual review and look at patterns across reviews. Do they see for a particular product a mismatch, for example, between review text and average star rating? Do they see patterns of many reviews posted in a very short period of time?

Fourth, instead of evaluating each review, consumers could outsource the assessment of reviews to algorithms or apps that they can find online. Fakespot and ReviewMeta are examples. You can upload an Amazon page and it will give you an assessment of the degree of fakery and adjust the star rating for you. If a product has a score of 4.8 on Amazon, maybe it falls to 4.2 once fake reviews have been removed.

Q: We're coming up to a big shopping season starting with Black Friday and Cyber Monday. You recently published a paper on practices that review platforms like Yelp and Tripadvisor can and do implement to fight fake reviews. What are these practices?

Wuyts: The problem of fakery is difficult for individual consumers to address, especially with new technology making fake reviews seem more and more real. So what we did in our research was look at players in the market whom we called guardians of trust. These guardians are the online review platforms, and we looked at the practices that they can and do use to weed out fake reviews and increase consumer trust.

The first practice is the actual frontline defense. That is, efforts using algorithms and human intervention to detect fake reviews. This happens on most review platforms that we observed, such as Yelp, TripAdvisor, Booking.com and Trustpilot.

The second practice is to expose the firms that post fake reviews or that pay reviewers to upload overly positive reviews for them. We observed fewer cases of exposure on platforms, but we found that the practice is a strong driver of consumer trust. About 25% of the platforms we observed use exposure. Yelp, for example, uploads alerts on the web page of firms when they notice that the firms have uploaded a lot of fake reviews.

A third practice is when the platform awards authentic, good reviewers with status symbols such as credibility badges. This happens on a number of platforms. We observed in a series of experiments that the implementation of status symbols also increased consumer trust in the platform. A main reason is that if you use this kind of status symbol, you increase competition for status among reviewers and you raise the bar a little bit for fakery.

The fourth mechanism that we looked at is what we would generally call socialization. In other words, allow for interaction between reviewers and consumers on a forum or in a Q&A setting. When a review platform enables interaction between reviewers and consumers, it raises the bar a little bit for fakery. This fourth mechanism also increases consumer trust in the platform.

A fifth practice that kind of completes the list, and something that may come to mind right away, is identity disclosure. What if reviewers simply identify themselves? They post a profile picture. They give their name. They share all kinds of other personal information. We found that this mechanism also serves to increase consumer trust. It serves as a cognitive shortcut for consumers to read reviews, and it also increases the persuasiveness of the message.

Now, one concern arises when we think of identity disclosure as a mechanism for review platforms, and that concern is privacy. Research has shown that as many as 85% or 87% of Americans can be uniquely identified based on their date of birth, their zip code and their gender. So when we ask reviewers to reveal their identities, we encounter issues with privacy, identity theft and so on. Plus, previous research has found that when reviewers have to reveal their identity, they are less likely to write really negative reviews out of fear of retaliation.

The good news is that when platforms use the other four mechanisms that I describe, identity disclosure does nothing anymore in terms of increasing consumer trust. We actually argue against identity disclosure and in favor of the other four mechanisms.

Q: Are you planning on doing any online shopping this upcoming season? If so, could you walk us through your personal process?

Wuyts: This study changed my behavior. If I want to buy something that is valuable, I would make the effort to run the web page through one of these online algorithms, like ReviewMeta or Fakespot. I would also go to one of these review websites like Trustpilot to see if that product is reviewed there and use those reviews to basically help or guide my decision-making processes and my purchasing behavior.

When you go to different review platforms, you can see how much the platforms communicate about what it is that they do to fight fakery. And for me as a consumer, this is something I didn't really do in the past but that would now be critical. If I want to use reviews to help me make choices, then I would go and verify what the platform is doing to weed out fake reviews.

Q: Is there anything else consumers should know as they shop online?

Wuyts: All these efforts are intended to improve the truth-to-lie ratio, so to speak. It is an illusion, I think, to eradicate fakery, especially with technology. Improving AI applications are making it increasingly difficult to detect fakery. Of course, algorithms that try to detect fakes are improving as well. But we're not really trying to minimize fakery to zero and entirely eradicate it. I think the real goal here is to make consumers aware that fakery exists and then also to make sure that mechanisms are in place to reduce fakery.

I would advise consumers to be vigilant when they read reviews. When they rely on reviews for making shopping decisions, they can identify reviews that are overly positive and short as fake. Perhaps they can use algorithms. But on top of that, I would recommend that they go and look for review platforms that communicate to consumers that they actively engage in all kinds of mechanisms to fight fake reviews.