Ratings and reviews bring an immense amount of value to a business. Almost all (95%) shoppers use ratings and reviews to evaluate or learn more about products, which influences their buying decisions and a business’ bottomline. According to our research, just one review on a product results in a 10% lift in orders, while 50 reviews causes a 30% lift in orders, and 100 reviews can lead to a 37% lift in orders.
Over the last few years, fake reviews have become more and more prominent in the mainstream lexicon and in the news. This is because unfortunately, along with the increasing value of ratings and reviews comes increasing interest from ill-intentioned parties, looking to take advantage of the solutions we have put in place to help consumers connect with brands and retailers. In order to ensure that both our clients and their customers are receiving real, trustworthy content on product pages, we have put in place best-in-class authenticity measures at every step of the way.
As part of our trust and privacy blog series, we interviewed Kevin Murphy, senior manager, authenticity. He oversees our authenticity team here at Bazaarvoice. Here’s what he had to say:
Can you walk us through Bazaarvoice’s authenticity processes?
“The authenticity team is responsible for making sure that the content that we are collecting for our clients isn’t fraudulent. There are a few ways that we do it. The real simple way to explain it is that when content is collected, it runs through some modeling and scoring tools. And based on the data that we get that’s associated with the content, the content gets scored and if it scores above a certain threshold, one of three things can happen.
It can be auto rejected, sent for manual review of it, or be approved if the risk of fraud is deemed low enough. The auto rejection process is based on rules that we write ourselves from patterns that we’ve identified, particular data signals or behaviors that we’ve figured out are risky.
Fraud tends to evolve. Once we’ve identified a pattern of fraud and stop it, people on the other side of things say ‘okay, time to change tactics.’ And then you start identifying new patterns. When we auto reject something, it’s rejected before it even goes to moderation. This also prevents the content from going live on a client’s site, which is the most important thing.
A big part of what we do is looking at the bigger picture of things. So when we’re looking at a particular piece of content, we’re really not just looking at the one piece, we’re looking at everything that is associated with it. Let’s say we receive a review from retailer X and all of the data signals that are in there link out to other reviews.
We go out and review these and start to look at the big picture of how all this stuff is connected. The whole point of fraud is a fraudster doesn’t want to put in any work or spend any money, but they want the returns. Once somebody sort of gets through one time, they’ll take advantage of that system flaw. So finding these patterns is really, really important. When we identify something as being fraud, a lot of times it connects out to other things, and we go and reject that content as well. So in some cases, content that may have been live already will get taken down.
The first time somebody commits fraud and that review comes through, it might look very, very clean. Because we have no history, we have no connections. There’s nothing on there, no red flag that will indicate that we need to look at this. But the more that person starts to do it, it starts to build up and you start to see these patterns. At some point, the risk score will be high enough to be flagged and then our team would go back and find all content associated with that user and investigate it.”
As a ratings and reviews provider, what responsibility does Bazaarvoice have to prevent fraudulent reviews?
“When I think about who it is that I’m protecting when I’m doing my job, I really think of it as being three entities. The most obvious one is that we want to protect our clients. Any type of fake reviews or fraudulent content has costs that can hit them. We’re also protecting ourselves. We are a service provider and the most important thing to a service provider is their reputation. We’re very upfront and outspoken about being the leader in authentic ratings and reviews. So we have a reputation that we need to protect and uphold. So in that sense, the more bad content that we’re enabling, that we’re allowing through, it’s really kind of an existential threat to us. So it’s very important from that aspect as well.
Then lastly, and this for me is the most important, is the consumers. I’ve been doing fraud prevention for more than 15 years. The whole idea of preventing fraud for me is predicated on the idea of being one of the good guys. Some of the people that are committing ratings reviews fraud, they’re genuinely bad people. They’re not just doing this sort of fraud. They’re doing identity theft, they’re using malware, they’re doing phishing, they’re doing credit card theft. It’s all connected. I think about consumers that are reading this content. One of the big benefits of our product is that it provides a lot of insights. So our clients get a lot of insights into what customers want and how they behave and all of that. If that information is based on fake content, that can drive clients to make very bad decisions.
And it can drive consumers to make very bad decisions if they are making decisions based on the bad content as well. Nobody likes to buy something and have it not meet their expectations. There are a lot of things that are riding on this content being true and genuine.”
Why is it important to badge incentivized content or syndicated content for clients and for consumers?
“First, there’s compliance issues. There are definitely mandates that we’re required to follow that come from the FTC and European regulatory agencies. Those mandates are based on the idea that consumers need to be armed with the right information. And sometimes the right information requires that you know a little bit of the backstory of a review.
So the badging, the attributions, content origin – all of that is really about making sure that the consumer has kind of a full story around the review, because there’s a lot of things that happen prior to that person writing the review that very much inform how they write the review.”
What would you say is the biggest differentiator between Bazaarvoice and our competitors when it comes to authenticity?
“A big thing is our fraud tools. I also think the idea of badging content and being very transparent are also differentiators. I think somebody can receive something for free and write a very honest review but again that transparency is really important.
We take this very seriously. My team in particular is very passionate about this. We have conversations with clients on a recurring basis where we discuss things that we’re seeing and why certain content was rejected. They send us content that they think needs to be rejected – there is definitely a feedback loop there.”
Want to to learn more about how Bazaarvoice helps our clients build authentic relationships with their customers? Connect with us here.