Eight in 10 Americans read online reviews before first purchases, yet the habit breeds decision paralysis. Experts: name preferences, skip the perfect pick.
James Hardaway reads the reviews before almost any purchase. A manufacturing manager in Texas, he checks Wirecutter and Consumer Reports, then browses Reddit for user complaints. He combs Rotten Tomatoes scores and press writeups before new films so he stays in tune with the cultural moment. Without that homework, he figures he might not be making the best use of his money.
The habit has not paid off as neatly as he hoped. He and his wife spent $450 on a top-rated stroller when their first child arrived and replaced it within months. They are not in love with their reviews-driven dishwasher or mattress, though the internet-approved noise machine works fine. He keeps wondering whether a different choice would have been better.
Every year, more products get a rating, he said. He may not need the best plunger, though something in his brain insists he does.
The research habit is widespread. Pew Research Center found in 2016 that eight in 10 Americans read online reviews before buying something for the first time. A 2026 Yelp survey found that an overwhelming majority of Gen Z consumers research on at least one platform before picking a restaurant or a gym.
Reviews tap the wisdom of the crowd, a genuinely useful invention. The volume is where the utility breaks down. Endless options and endless opinions leave shoppers less satisfied with what they pick and strip spontaneity from a simple pizza order.
The dinner example is familiar. You decide on pizza, start browsing local options, and 30 minutes later you have read dozens of reviews across 15 restaurants without a decision. One place is probably better than the rest. Any pizza is probably fine. Somewhere in that gap a Tuesday night disappears.
That drive to find the best carries a cost. "Once you're out for the best, the path to misery is pretty straight," said Barry Schwartz, professor emeritus in social theory and social action at Swarthmore College and author of "The Paradox of Choice."
Michael Luca, a professor at Johns Hopkins Carey Business School who studies online ratings, says the question is when to demand the best option and when to accept something that merely works.
The platforms reinforce the instinct. Google surfaces stars before websites, and Amazon shows reviews before the price. AI search engines now summarize the consensus before a user sees a link. Small businesses live and die by the algorithms and online rating systems.
AI systems now boil hundreds of opinions into a few sentences for the major review platforms. Research shows the summaries lift purchase intent for electronics and appliances. For restaurants and hotels, they do not move the needle. Power users who like to comb through the original ratings get little from a compressed overview.
Accuracy is another problem. Summaries can flatten commentary into a generic package or overweight a small number of negative comments in a bid to seem fair. They can also get the facts wrong. Small business owners on Reddit complain that the AI overviews include extra-negative or incorrect content.
AI could help if it tailors information to a person's preferences, Luca said. A one-size-fits-all summary, he said, does not get anyone closer to the product that fits.
Doug Straton, chief marketing officer at Bazaarvoice, a company that helps brands collect and display ratings, says the problem starts with what the crawlers read. They look at the volume and recency of reviews when deciding which products to mention, not the context behind individual opinions. A reviewer who dislikes cucumber may rate a bodywash poorly, and the AI overview tells readers only that the product smells bad. "Because you're superficially scraping the sites, you don't have this depth of data," Straton said.
His company works with AI firms to give more weight to Bazaarvoice's own community reviews, which he said are more thorough than random web scrapes.
Fakery is the second layer. AI tools make it easier than ever to spin up reviews and harder than ever for consumers to tell what is real.
Kay Dean, a former federal criminal investigator who founded Fake Review Watch, said fake reviews have been a "massive" problem for years. Platforms are not aggressive enough about policing them, she said. "They sweep it under the carpet."
Businesses barter and trade reviews, and some recruit ringers through Facebook groups. Review brokers sell ratings in bulk, Dean said. "There's a whole cottage industry of review providers." She described the market for fake reviews as open and flagrant.
The Federal Trade Commission responded in 2024 with a rule banning false ratings and testimonials and imposing financial penalties on violators. The rule also bars incentivizing reviews with a specific sentiment. Dean called the rule a good step. Review platforms escape liability, she said, because federal law generally shields online companies from responsibility for content users post.
No one is arguing for a return to a reviews-free society. Sheena Iyengar, a professor at Columbia Business School who studies choice and innovation, says clarity on preferences is the fix. Looking for the "best" bar or the "best" gym leads to a long search and disappointment, she said. A search for a quiet spot with decent espresso martinis, or a gym with plenty of weights and spin classes, tends to end better. Most situations do not warrant that level of inspection, she said. Big decisions like booking a vacation deserve the slow read; the rest do not.
"You make a lot of choices, and most of them are irrelevant and unmemorable," she said. Most of life's choices, she added, are better made on autopilot, saving cognitive energy for the ones that matter.
Hardaway still checks reviews, and probably always will. He has learned the online chatter is not gospel and that research does not guarantee a good outcome. Sometimes the best stroller is not the one the gurus recommend. It is the one a family actually uses.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.