How Memory Affects the Language of Fake Reviews
Have you ever read a review and thought, wait, that sounds like it was written by the company… ?
There is research into this actually, and it turns out that fake reviews tend to share similar language construction:
Fraudulent reviews tend to be more extreme than authenticated reviews, regardless of review valence. Using an experimental design Kronrod et al. (2017) instructed participants to write reviews for a hotel stay that they had or had not actually experienced. They found that fictitious reviews generally used fewer verbs in the past tense, fewer unique words, and more abstract language due to the lack of actual experiences and concrete memories.
Detecting Fictitious Consumer Reviews
Last week I wrote on how Episodic Memory and Foresight are major influences in our purchase decision-making process. We must be able to imagine ourselves using a product, before we are going to buy it. Just knowing the functions and features is not enough.
Kronrod and her colleagues uncovered something similar.
In their original paper titled Been There, Done That: How Episodic and Semantic Memory Affects the Language of Authentic and Fictitious Reviews, Kronrod explains how genuine reviews draw from episodic memories. Concrete memories. With unique words. Details and situational descriptions.
Lacking real memories, fake reviewers use vague and abstract language. They don’t speak in past tense… because they are thinking in present tense or in a future (therefore hypothetical) tense.
I would recommend vs. I recommended them to…
Experienced fake review writers know to add in specifics and speak in past tense, but you average free-product-for-review taker is not going to know how to rewire their brain so quickly.
I would… is the verb usage we use for hypothetical situations. Like… I would wear a raincoat if it starts to rain. I would get a ticket if the police clock me going 50 in a 30. Future possibilities.
Past tense usage can be more specific.. I wore a raincoat this morning because it was raining when I went to work… or Officer Barbrady gave me a ticket for going 47 in a 30.
It is not that fake reviewers can’t dream up specific details, but our brains don’t worry like that. Unless we have practice, most of us are not pathological enough to switch so fast.
Here is a screenshot from the website of a shoe brand. Notice the verb tenses…
This is from the Cariuma website. Earlier this year during a customer interview, a woman named Raquel shopped this site and got suspicious when she made it to the reviews on a certain shoe she liked.
Right off the bat, 100% 5-star reviews. All 38 of them. As a shopping aficionado, she instantly thought it was off. She didn’t read the reviews, and instead googled Cariuma to hear from 3rd party review sites.
It was an immediate ding on her trust in this brand.
In a 2016 study, researchers concluded that:
Legitimate reviews show some consumer consensus and tend to be unimodal, whereas the distribution of fake reviews is bimodal, with greater frequencies at both extremes (e.g., 1-and 5-stars)
Detecting Fictitious Consumer Reviews
We are not as good of liars as we think we are.
A couple of years ago a friend called me to ask a favor. His little sister published her first book on Amazon. He wanted me to buy it and give her a review. So I did.
I read a couple of pages, and it was good. Some part of me thought I would feel better about leaving a review if I at least read some of it. I left a 5-star review and wished her the best of luck.
Was that a fake review? Technically, yes.
Was it unethical? I don’t think so…
Amazon is like that, and reviews are important to ranking. I was helping a friend and a budding author. She was so excited when she read her first review.
I don’t think Amazon wants to stop me from leaving that review… what they want us to believe they are trying to stop is this lady:
I used to make purchases and leave a review on Amazon and in turn for the positive review, the seller would provide a rebate, discount, or free product. Amazon felt that my review activity was odd. They removed every review I did, (hundred+) banned me from leaving reviews and was also stuck in absorbing the costs associated with the purchase of the products that I purchased that I could no longer review due to the ban.
Question is though, people that review or test products, shouldn’t they be fairly compensated for a “job” done? Why ban them? My reviews were authentic and true.
Tracy
Tracy said this publicly in a comment someone’s post. She believes she is doing us all a service… and should be fairly compensated for her hard work.
Thing is why we can’t have nice things, Tracy!
Fake reviews are a feature rather than a bug in ecommerce.
Marketplaces like Amazon know this. This reported 900 million is performative… it’s just as fake as their commitment to stamping out fraudulent reviews.
This is an abridged post from Saoud Khalifah, the CEO of Fakespot:
We've looked at the most fake reviewed categories on Amazon after the record breaking Black Friday/Cyber Monday craziness. Here are the top 5:
Watches = 45%
Trees = 45% (tied with watches)
Smartwatch Bands = 45.3%
Pullovers/Sweaters = 59.5%
Slippers = 71%
This should not have been surprising but it was. Try searching this category on Amazon and take a look at the results. No name brands with tens of thousands of ratings (see below image).
Those are some hefty numbers. If Fakespot is right, these operations behind these reviews are massive.
Here is an excerpt from a story about one PhD student investigated these review operations for his
Agents are trained on how to recruit reviewers (referred to as “Jennies” by the agents). Tips for recruiting folks on Instagram, for example, suggest following hashtags like #Amazonreviews, as well as experimenting to find the best time to post about products. Agents are shown an example of an attractive prospect or “Virgin Jenny,” an existing reviewer profile with a single review on it.
Read the full article on Wired, it goes in-depth…
Brand websites… marketplaces… and of course local listings.
Nowhere is safe from these operations:
I was solicited from a company in India that asked me to give them $300 for 50 reviews. I wrote a heavy worded email back and warned them that doing this type of thing is very poor business and will lead to prosecution.
Brad
I’d like to believe that local business people are honest like Brad and won’t engage with these outfits, but…
Google reviews are a constant source of complaints for small businesses. Google’s algorithm frequently flag and remove legitimate reviews.Some companies have marketing agencies who post/solicit fake reviews.
In the past year, has Google has posted 47 five star reviews for my law firm. I know for a fact that other reviews have been left, but they were not posted.
One of my competitors has received more reviews THIS WEEK than my law firm has in a year….
I’m not sure what to do, other than to keep doing good work and asking for reviews from folks that I’ve helped. It’s unethical to have a vendor posting fake reviews. What goes around, comes around. My hope is that Google’s algos improve and penalize the companies that are creating countless fake reviews.Hunter
Local businesses and service are just as susceptible to fake reviews as the rest of the internet.
Customers, like Raquel, and like you, already know to take online reviews with a grain of salt. Sometimes they are legit. Sometimes they are fake. Sometimes they are from a disgruntled customer trying to get even.
Regardless, they play an important role in ecommerce and in a customer’s decision marketing process.
I’d like to hear your thoughts and your experiences with reviews…
If we are in for turbulent times like they keep saying, are these “hacks” going to get worse or will businesses find more creative ways to appeal to new customers?
What’s your take?
Do you give much weight to online reviews?
Are you good at spotting fake reviews?
Have you ever given or been asked to give a fake review?