Customer reviews wield ever more power over purchasing decisions as online shopping grows and with it consumers’ tendency to publicly share feedback.
If you’re shopping for a dog bed on Amazon, an eye cream on Sephora, a dishwasher at Home Depot, or trying to decide which hotel to stay in with your family in Aruba, chances are you’ll read through a few customer reviews before clicking “add to cart.”
Virtually all websites selling products or services invite consumers to leave product feedback in the hope the comments will compel other shoppers to buy.
Nearly all online shoppers — 99.75% — read reviews at least sometimes, while 9 in 10 do so always or regularly, according to a 2023 survey by Power Reviews. It also found that more than three-quarters of shoppers specifically seek out websites with ratings and reviews.
In their ideal form, reviews provide other consumers with a first-hand experience and enough information to determine if a product will meet their needs. But would you expect a customer to read more than 8,000 reviews to determine the best place for a winter vacation?
At a certain point, the endless scroll of consumer reviews isn’t helpful.
It’s a challenge customer experience leaders face: While shoppers want to read reviews to inform purchases, they can also present seemingly unending choices and information and overwhelm to the point of decision paralysis.
Nearly three-quarters of consumers walked away from purchases simply because they felt overwhelmed, according to an April Accenture report. Research shows there is a limit to how many reviews are useful.
Too much information
Companies like Amazon realized the volume of reviews on some products stopped being helpful to customers. The mega-retailer started featuring “helpful reviews.” If a shopper believes a review is useful, they can click the “helpful” button below a review, which boosts its ranking among the top reviews shown on the product detail page.
In a perfect world, a customer is well informed after reading reviews and is satisfied with the product they purchased.
“You should expect subsequent shoppers to be happy with their purchase, on average,” said Wael Jabr, assistant professor of supply chain and information systems at Penn State University.
In a study of products and product reviews on Amazon published in 2022, Jabr found that those “helpful reviews” can lead to higher sales and improved customer satisfaction. But the number of “helpful reviews” presented matters.
Jabr and his co-author compared products with three “helpful” reviews to ones with more than three across four categories — automotive, grocery and gourmet food, health and personal care, and home and kitchen. When products featured more reviews, the impact of the reviews diminished.
“If there are seven reviews, and there’s a variation in reviews, it could be problematic,” Jabr said. “There’s cognitive dissonance — you reach a threshold of information, after which it becomes harder to understand.”
The bottom line: humans simply aren’t great at distilling contradictory information.
Han-fen Hu, associate professor at the University of Nevada, Las Vegas, in 2019 published research that looked at whether online product reviews help or hinder the decision process. She found that too much information in the form of product reviews can negatively impact other shoppers' experience.
The relationship between information volume and overload is not linear. Rather, it's an inverted U shape. When consumers are overloaded by information, they’re less satisfied with their own purchase decisions, said Hu.
The AI summary
As a way to manage a glut of reviews that can overwhelm shoppers, many are following Amazon in harnessing the power of AI to summarize consumer reviews. Summarization can make it easier for shoppers to get the general sentiment without having to slog through pages of other customers’ feedback.
Best Buy recently added the ability to filter customer reviews of products by themes, like battery life. And TripAdvisor this year added an AI-powered feature that provides summaries of recent traveler reviews on hotels based on user-identified quality attributes.
“Consumers are likely to rely more and more on these summaries to get the information they need quickly, choosing to dive deeper into reviews if a specific facet of the product they are interested in is not included in the summary,” said Steven Bailey, customer and growth leader for the consumer sector at EY.
The next step for AI might be to remove the need for consumer reviews altogether. As uses for generative AI proliferate, companies can use these technologies as a new form of “influencer” and adviser, said Jatinder Singh, global head of data and AI at Accenture Song.
If someone is shopping for something specific — say, a bicycle for their 7-year-old child — and needs more information before making the purchase, generative AI advisers can “fast track the discovery process, showcase the options, and deliver expert, personalized advice,” according to Accenture’s April report.
In the next iteration, generative AI acts as an “agent” or personal assistant that’s capable of taking complex decisions so the consumer doesn’t have to make as much of an effort. “They're starting to understand what’s important for you and give you advice,” Singh said.
