It’s Saturday night, you’re somewhere downtown, and the bathroom line is long. Like, 15 people long. Some are staring at their phone, some are glancing out at the dance floor wistfully, like kids in detention watching recess. The woman in front of you is wearing the most incredible red lipstick you have ever seen, a vibrant red that skews orange when the strobe light hits. Not too matte, not too shiny. Is it Dragon Girl? You don’t know, but you have to ask.
So you do. What unfolds is natural: The two of you share a tenuous but palpable bond, right there, in the bathroom line, over that red lipstick. You were looking for something, somebody else discovered it, and you found that person. Chatting about lipstick turns into chatting about life. Maybe you become friends, or maybe you never see each other again. But you got the shade name. And the horizon of your world spreads a bit broader.
Until you find out Dragon Girl wasn’t a girl at all, but an influencer planted there to tug on that thread of beauty curiosity. Or a robot programmed to spew praise at anybody it registers as a human woman age 28 to 40 with an annual income of $50,000 or more in the American Midwest. Or a lipstick brand masquerading as a beauty reviewer on a large retailer’s website. “The perfect red-orange,” she says. “Just the right amount of matte-to- shine ratio!”
To read online beauty reviews is to negotiate such potential minefields. The question is not whether reviews can be faked. We know they can because they have been. Last year, a Reddit user claiming to have worked at skin-care brand Sunday Riley posted what appeared to be an email with the subject line “Homework Time —Sephora.com Reviews.” It walked employees through a step-by-step process on how to hide their IP address (an easy maneuver to make it appear that you’re using your computer in, say, Japan, when in fact you’re at home in Hoboken), leave a minimum of three reviews for one of the brand’s newest products, and, perhaps, mention specific qualities of the product that had been maligned in other legitimate reviews.
Within 48 hours, Sunday Riley confirmed the veracity of the email, saying it was indeed sent by a former employee and they would be “making an effort to bring more transparency” to their customers. They added that it would have been “physically impossible” for them to have had employees post even a fraction of the thousands of online reviews of their products. “We committed in 2018 to publicly releasing audits of our business practices and ethics,” the brand told Allure in an email. Now employees are discouraged from leaving reviews, and if they choose to do so anyway, they are obligated to disclose their employment with the brand.
Interviews with beauty entrepreneurs, site managers, consumer psychologists, as well as influencer agencies tied to the product-review industrial complex confirm the fears of beauty shoppers everywhere: The word-of-mouth recommendations we rely on when filling our vanities aren’t always trustworthy. “We were told, ‘Okay, everyone’s tried the new product; go write a review on our website,’” says Sally Smith*, an employee at another skin-care start-up.
The behavior isn’t surprising considering the stakes: The global cosmetics industry is expected to be worth about $800 billion by 2023. And recent data shows that 77 percent of online shoppers read product reviews for more than half the products they purchase. Besides having had a positive experience with the product themselves, seeing a large number of positive reviews makes digital shoppers trust a brand. A lot of money is on the line. And brands aren’t willing to leave it all to chance.
That dovetails with a beauty-consuming public more eager than ever for insights about skin care, hair care, and makeup. “In [this] space, people are hungry — starving, in fact — for information,” says Kit Yarrow, a consumer psychologist and professor at Golden Gate University. “They understand that there is constant innovation in beauty [and] that also means constant learning.” Personal recommendations are a way for the masses to keep up with it all.
Reviews move product. Period. “They’re heuristics—mental shortcuts that allow us to simplify choices — that help in making purchase decisions. That’s why we like and use them,” says Peter Noel Murray, a New York City-based consumer psychologist. “But especially in beauty and skin care, it isn’t about functionality. It is about the experience. What emotional end- benefits does it provide to me?” How exactly can that be qualified online? It can’t, really. But it can be quantified.
“You want to always have a 4.5-star rating,” says Justin Jackson*, who has handled reviews for multiple start-ups’ websites, including a skin-care and cosmetics company known for its millennial-cool branding. Jackson, as well as many others interviewed for this story, asked that we not use his real name. “If you go on someone’s website and they have 500 five-star reviews, you’re going to be like, ‘That’s fake.’” On the other hand: “If it’s 500 and it’s 4.5, you’re going to be like, ‘Huh, that means people really like it and maybe for one or two people it just didn’t work out.”
That perception doesn’t necessarily come without some behind-the-scenes sorcery. “If you’re managing a site, it is possible to edit the reviews that come in, and you can kind of…” he trails off, inviting me to finish his thought. In the star-rating realm, though, the floor is scarily close to the ceiling: “Anything below a 3.5 and your product is done, says Kevin James Bennett, a consultant who has worked with multiple brands that sold their wares through Sephora. “The perception is 3.5 is a full star and a half below a 5. People are not happy. I mean, be honest — when you look at something and it has a 3.5, you keep looking. Boomers are said to have a five-second attention span when we get to a website. Millennials have four seconds; Gen Z has three. If you don’t make a good impression in three seconds, they’re gone.”
Sales aren’t all a brand stands to lose: “If you fall below 3.5, your product pretty much falls into the danger zone,” Smith says. More specifically, says an independent beauty consultant who agreed to share details anonymously, the danger of possibly being RTV’d: returned to the vendor. “They don’t play,” he says. “This is what destroys companies. If [a product] doesn’t sell, they have the right to return to the vendor at will. It could bankrupt an indie company that doesn’t have the capital to absorb the deficit.” (Sephora declined to provide an interview for this story after we sent questions regarding the review process on their site.)
Stellar reviews aren’t enough — brands are pressed to have as many of them as possible. Because we don’t just want feedback; we want copious amounts of feedback. A study published in Psychological Science found that when given a choice between two equally low-rated products, people routinely chose the one with a higher number of reviews — in other words, the “more popular” product (and the one statistically more likely to be of low quality). Retailers understand this. “During your brand review twice a year [you’ll be told], ‘Your star rating is above average for this category, but you don’t have a strong volume of reviews,’” says Danielle Davis*, a brand founder who has sold products through Sephora and asked us not to use her real name. “They push brands to have reviews — and a lot of them–which means some brands do questionable things to get them.” (More on those tactics later.)
Of course, Sephora isn’t alone in its supposed demand that brands pony up reviews. A brand founder seeking distribution with luxury beauty retailer Violet Grey disclosed to Allure that the company asked outright for the founder’s customers to review products to post on their site at launch. In an email, Violet Grey denied suggesting its brands accrue reviews. Instead, the company reaffirmed its vetting practice, which includes sending products to its 100-plus- member expert committee, whose positive feedback is crucial to Violet Grey’s decision to carry a new product. The company discloses that these members are also Violet Grey customers.
In our broader “cancel culture,” you’d be hard-pressed to find a brand that is still fabricating its own reviews. Such a public misstep isn’t worth the risk. And yet, Bennett says, “I don’t think anyone’s cleaned up their act completely — they’re just being more careful.”
Enter two strategies that fall firmly into a gray area. They were mentioned so often in interviews for this piece that I suspect they generate the bulk of reviews you’ll find on major retailers’ websites. The first is “seeding” products — sending them to existing customers (usually through loyalty programs) or via third-party services like Power Reviews, Influenster, and Bazaarvoice. These companies — born of our reviews-driven shopping culture — will get your product to women in your target demographic, who are then highly encouraged to leave reviews. It’s easy to carry a positive bias into a review, though, if it was received for free — a gift by definition — instead of having required some investment. (Full disclosure: As an editor at Allure, I am the recipient of the seeded products. But Allure’s editorial policy is that staff may accept free goods as long as their acceptance does not compromise or guarantee editorial coverage of the free goods. We never solicit gifts, and we also purchase products for review.)
Brands adamantly deny paying for reviews, but an expensive seeding or gifting campaign, with the goal of generating positive buzz for a product, can feel pretty similar. It gives brands with deeper pockets a better shot at stacking their page with a plethora of reviews, thereby inspiring you, dear shopper, to pull out your credit card with confidence.
The second strategy, incentivizing, can be even more enticing for a potential reviewer. Though it’s illegal to incentivize a review, according to the FTC — with free product, cash, 15% off coupon while supplies last— unless disclosed, beauty company employees I spoke with said brands frequently offer impressive rewards to reviewers, such as credit to use on the brand’s website or discounts, on top of free “seeded” product.
So yes, deep pockets certainly have their advantages in the review game. But sources at conglomerate- owned brands seem to think that indie brands have their own edge: less accountability from a more established corporate structure — and those pesky shareholders. “It’s easier for smaller brands to get away with generating fake reviews than it is for bigger brands, but we joke about it, saying things like ‘We should be writing reviews and refreshing the page,’” says Aaron Adams*, a copy manager at a skin-care brand owned by a prominent parent company.
Direct-to-consumer brands, for their part, have found a convenient loophole that allows them to get up to thousands of reviews per product while keeping up appearances: “ambassadors” who work on commission. “A lot of the reviews we had were from ambassadors,” says Josh Johnson*, a former social media manager for a direct-to-consumer brand. “We never sent out emails that said, ‘We want you to write a review.’ Ambassadors are really passionate about these products.” And why wouldn’t they be when they’re getting a share of the spoils?
Positive reviews aren’t the only ones that can be gamed. Another brand founder, Beth Booker*, whose line was available on a television-based shopping network’s website, once suspected a competitor had left dozens of bad reviews for Booker’s new foundation on that site. A cursory username search revealed that the same people were leaving five-star reviews for that competitor’s product and two-star reviews for Booker’s.
For better or worse, brands say it’s extremely difficult to get product reviews taken down even if they suspect foul play. In the case of the shopping-network reviews, the brand was too big of a cash cow to receive more than a slap on the wrist, or so Booker thinks. In most cases, though, a third-party retailer’s reluctance to scrub negative reviews is a point in their favor for being a trusted source. Says Lucy Lake*, a social media manager at a skin-care brand: “[In-house] we flag ones that seem off, but unless it’s inflammatory, Sephora doesn’t remove it. You’re seeing everything.”
Sowing The Seeds:
OK, so for those keeping tabs at home, we have the ideal star rating (a 4.5), a nonstarter rating (a 3.4 or below), and a desired volume (a lot). But there’s also the right timing early and often. Certain retailers expect beauty companies — and possibly their own websites — to seed products before they launch so that there are dozens of reviews waiting for would-be customers when they arrive at that product page on launch day. “We always aimed for around 100 reviews at launch,” Smith says. “That seemed like a good sweet spot.” At first, it makes no sense at all.
What kind of sucker is going to believe reviews from people who got some sort of VIP access in the first place? On the other hand, it makes perfect sense. Think of how you’d feel if you met someone new who seemed promising, and they told you that you were their very first friend. You’d wonder if everyone knew something was up with them—or worry you’d discover so yourself.
And one more thing: “Recency matters too,” says Laura Brinker, senior vice president of marketing at Influenster, one of those companies that specialize in getting beauty products in the hands of potential reviewers. “Most people think that if the review is more than three months old, it’s not particularly credible.” As consumer psychologist Murray explains it, “the currency of the internet is immediacy.” But don’t worry! Influenster offers a package to help. For the right price, you can seed products over months or even years to ensure a steady stream of reviews for existing products. (The brand emphasizes that seeded reviews only comprise 2% of their overall reviews, and that the vast majority are written after product purchase.)
With such behind-the-scenes machinations, a cottage industry has emerged: web-based services, such as FakeSpot, that help you weed out fake reviews. Just paste in a product page, and AI-based algorithms will determine which reviews come from bots or influencers, which it deems less credible. FakeSpot currently analyzes reviews from the likes of Sephora and Amazon — in June, it declared that almost 25% percent of the latter’s beauty reviews are untrustworthy. Let that sink in: According to FakeSpot, about a quarter of Amazon’s beauty reviews are fake, with cosmetics, anti-aging, and hair-loss product reviews harboring the biggest phonies. A spokesperson from Amazon declined to confirm a percentage or share figures on untrustworthy reviews. Last year, the company estimated this figure at less than 1 percent, but no explanation was provided for the gulf in estimates. However, the company says about 90 percent of inauthentic reviews are computer-generated, and the e-tailer uses a combination of machine learning and investigative teams to weed out suspect posts. ” Even one inauthentic review is one too many,” says the spokesperson.
At press time FakeSpot had yet to look at ulta.com. We spoke to many brands that sell at Ulta Beauty, and while they all say reviews are certainly important to that business — and that they have seeded products before they’ve launched there—they didn’t feel any serious pressure to hit certain numbers.
Almost every retailer has its own methods to help bar fake reviews from its site, or at least designate which ones might be more authentic than others. Ulta sends emails to buyers after they make a purchase so they can post the review as a “verified buyer,” and Sephora has an honor system that asks reviewers to indicate whether or not they got free product. (Many, but not all, brands also ask for this disclosure from consumers who review their stuff, but it is a difficult request to enforce.) What’s more, at least one source told us they’ve heard that if you upload a high volume of reviews on one product from the same IP address, “Sephora will flag it.”
So no, this world is not the Wild West completely. But still: As long as reviews exist, brands will push—and be pushed—to get good ones. “There is a huge pressure for every single product we launch,” Adams says, a little defeatedly. “At the end of the day, it’s ‘anything to sell products.’”
There is hope, though: Not all of us are that easily sold. “I’ve seen shoppers check two or three retail websites, watch videos, and google bloggers to purchase an eye cream,” says Yarrow. Because yes, they want to know if the color payoff is good and how your skin looked the next day. But also: “We humans need support,” says Murray. We want to connect. In a world where choices are vast, we want to know our decisions are right and well-informed, so we’ve come to look at the stars (3.5 and above). Whether it’s the right place or not.