Online reviews shaped the internet as we know it. Now they might be in danger.

In a primitive world before George W. Bush’s presidency, iPhones, Olivia Rodrigo‘s birth, and the release of the seminal Romy and Michele’s High School Reunion, Eric Goldman just wanted to find a good vegan restaurant.

We all know what we’d do in our post-Romy and Michele world: Within seconds, you’d practically be on autopilot, scrolling through troves of online reviews, scanning for vital signs. (5 out of 5 stars? 1,000 reviews? Done.)

But 2021 might just be the time to revisit that bygone world before the online review ecosystem we know now — the one that Goldman, a Santa Clara law professor who served as general counsel for one of the web’s first review platforms, Epinions.com, found himself in decades ago.

Why now? The pandemic, and the boom in online shopping that it brought on, put a spotlight on online reviews, as well as the duplicity that sometimes surrounds them, while chatter of either axing or tweaking the bill that allowed online reviews to exist in the first place has ramped up too.

Together, those factors invite a perilous question: What would we stand to lose if we lost online reviews?

That’s not going to happen tomorrow — or, potentially, ever — but in order to even think through the question, it’s worth going back to the beginning, and that’s where Goldman and his search for a vegan restaurant comes in.

“The next big thing”

According to Goldman, the search for a good vegan restaurant used to be something of an ordeal. Up until the late 1990s, there were only really two options, neither of them ideal: You could turn to professional reviews, weighty tomes like Zagat guides, or you could ask friends, who might be picky eaters or uninformed about vegan cuisine. “There’s really no offline analogy to [online reviews],” he explains.

The idea of an online hub of information, powered by user-generated content, about what to buy and where to buy it was unheard of in the early ’90s—particularly since the Web 1.0 internet hadn’t yet prepared users for a barrage of opinions about everything all the time.

But along came Epinions, where Goldman served as general counsel, a startup so innovative that even job applicants were kept in the dark about what the company actually did. The idea was achingly simple in retrospect: “Zagat-for-everything.” The logic was straightforward: People bought stuff on the internet, and they needed a way to figure out what was legit versus what would arrive on their doorstep looking just a little different than the image online. The key, though, was that they’d be getting this insight from people just like them.

Vetting products online was a genuine concern for consumers as e-commerce picked up steam, according to Dr. Joseph Reagle, a Northeastern University communication studies professor who studied online reviews in past work about online comment sections. “​​Back in the ’90s, most of the people selling things [online] were certain niche merchants, so even Amazon was a bookseller,” he explains.

While people were initially wary about giving websites their credit card info, as the public got more comfortable with the concept of online shopping, the key security question changed, too. It was no longer a question of the merchant’s reputability, but rather, Reagle says, “are the products worthwhile, even if I trust the merchant? And for that, people introduced reviews.”

Amazon already established the basic concept of consumer reviews for books before Epinions emerged. Yet Goldman says it was in the wake of Amazon “showing the value” of this type of review that an emergent class of sites just for reviews popped up as well. Unlike Amazon, he adds, they “weren’t beholden to a particular retailer,” which meant these sites could cover “a wider range of goods and services.”

Epinions was one of the first, but it joined a cadre of other largely forgotten websites, like Deja, a discussion group site that eventually focused on consumer reviews, and RateItAll. Like so many others, they’re vestiges of the dot-com boom, a time in which the web’s futuristic sheen meant there was enough cash flow to allow for experimentation.

When it came to consumer reviews, “the late 1990s was a really interesting innovation period,” Goldman explains. “Was there a business model that was sustainable or not? We weren’t sure.” He adds that not only was it unclear “where consumer reviews fit into the firmament of the internet” in the late ’90s, but also “if they were one of these crazy ideas that was going to fail, or if they were the next big thing.”

The bill that launched billions of opinions

Of course, we now know consumer reviews were, indeed, the next big thing — and Goldman points to a single foundation for that: Section 230 of the Communications Decency Act of 1996, which protects “internet computer services” from legal liability for content users post on their site.

“The timing wasn’t accidental between Section 230 being placed on the books and then a service like Amazon experimenting with consumer reviews, and then standalone services like Epinions coming to the fore,” Goldman says. “And the reason is that consumer reviews are just a giant liability trap.”

Without explicit protection for online speech, plenty of businesses would be tempted to claim negative statements (“this pizza is BAD”) were defamatory or otherwise genuinely harmful, Goldman says, and “no business was willing to experiment with giving amateurs the power to criticize businesses, unless they could avoid liability for it.” Reagle agrees that Section 230 was integral in the creation of new platforms offering reviews. Indeed, after Section 230, more and more of the standalone consumer sites we know today started to emerge as well: TripAdvisor in 2000, Yelp in 2004, Goodreads in 2006.

“Was there a business model that was sustainable or not? We weren’t sure”

Meanwhile, many of the absolute firsts — the Epinions and Dejas of the world — were later acquired by bigger tech companies, or otherwise saw their general ranking systems folded into new companies that cropped up. Reagle notes that once reviews became “important properties that people wanted to spend a lot of money on,” that was when we also saw a string of acquisitions: Google bought Deja, Amazon bought Goodreads. (Notably, Google and Yahoo tried to buy Yelp, to no avail.)

In relatively short succession, an entire online review ecosystem emerged, with companies that were able to acquire others quickly surfacing as the power players. And once online reviews became ubiquitous, in turn, it became next to impossible for businesses to ignore them. “Everyone, at some point, decided they needed to start doing online reviews, or they would be left behind,” Reagle explains.

Notably, Goldman notes, we haven’t seen many changes with respect to which platforms we turn to. “There was a lot of experimentation that took place between 2002 and 2010,” he says of online review platforms. “When was the last time we had an entrant of a new standalone consumer review site? There might be some really niche-y services that are cropping up, but for the most part, the major services are the same today as they were 10 years ago.”

And that also leads us to where we are today.

What happens to online reviews now?

We’re still living in a world altered by this late ’90s, early ’00s explosion of online reviews — even if the major platforms have largely remained the same.

Alongside the growth of online reviews, fake reviews have proliferated, as well as review bombing, the phenomenon in which users dump heaps of negative reviews on businesses or creative works they’re looking to critique. Saoud Khalifah, the founder and CEO of Fakespot, an app designed to spot fake Amazon reviews that was removed from Apple’s app store after Amazon reached out to Apple, says the explosive growth of online reviews “has opened up this Pandora’s box,” incentivizing legitimate businesses to gain and rig reviews.

There aren’t concrete stats on the absolute total number of fake reviews across all of the major review platforms, Kay Dean, a former criminal investigator now devoted to tracking fake reviews, points out. “But for context, Google said that it removed or blocked 55 million reviews in 2020, and Trustpilot said that [it] removed 2.2 million fake reviews in 2020,” Dean explains, adding that she found “tons” of additional fake reviews on those platforms in the same timeframe.

“Consumer reviews are just a giant liability trap”

Regardless, we know fake reviews and review bombing abounds, and it all poses a massive moderation conundrum. The protections afforded by Section 230 — a.k.a. the very reason online reviews emerged when they did — are a big part of that conversation. As of late, though, enthusiasm for Section 230 has soured, and that could have implications for review platforms if protections are rolled back.

Former President Trump was a strident Section 230 critic towards the end of his presidency when his social media posts were being subjected to moderation over their inaccuracy. Other Republicans have pushed against Section 230 in hearings about social media platforms, particularly the ways in which Section 230 shields them from liability. Across the aisle, Democrats have increasingly called for Section 230 reform as a potential way of cutting down on harassment and discrimination, and a more general response to the largely unfettered growth of big tech.

And following Facebook whistleblower Frances Haugen‘s explosive Congressional testimonies, calls for Section 230 reform have reached something of a fever pitch.

“This is the conundrum. It’s always easy to see things in hindsight,” Reagle says. “One of the reasons why I think the Communication Decency Act and Section 230 was really a good thing at the start, is you don’t want to unnecessarily stifle the development of a new marketplace.” He adds that while Section 230 served a crucial function in the internet’s foundation, he hopes “maybe now we can think about this a little bit more carefully, when we think towards the future.”

Others, like Goldman, remain staunch Section 230 defenders. If we got rid of Section 230, he points out, consumer reviews “as a class of goods” would vanish instantaneously. The calculation, then, becomes a question of how much we value those reviews.

“As someone who remembers what life was like before consumer reviews, I will tell you, vendors work so much harder today to satisfy their customers than they used to in the old days,” he explains. “Now the power of consumer reviews to be held accountable publicly has meant that businesses go above and beyond to try and satisfy their customers.”

“I still think consumer reviews are essential to the overall well-functioning marketplace, that they solve an information problem that we need solved,” he continues. “We don’t want to go back to relying only on professional reviews; we don’t want to go back to relying upon word of mouth, and some idiosyncratic views from our friends.”

Khalifah and Dean, however, think we’ve largely been trained to crave online reviews, and that we could often rely on them less, particularly in the face of the fakery that’s currently proliferating.

For now, with some review platforms rolling out more moderation efforts while others stand idly by, most of us have found some workarounds for fake reviews, whether reading reviews with a healthy dose (or abundance) of skepticism, or checking multiple sources for reviews when possible. Others, like Dean, however, wouldn’t mind a full slide back into the past, eschewing online reviews almost entirely — at least until tech companies clean up their act. She wants to see Section 230 reform, but more than anything, wider accountability for tech companies that fail to regulate fake review trading on their platforms.

“These tech companies know what’s going on, and they do far too little to police their platforms because they don’t have much incentive to, in the current environment,” she says. “I feel like the Erin Brockovich of the online review, because that’s how I felt screaming from the rooftops.”