Navigating authenticity in online product reviews
Research by Lingnan University in Hong Kong is providing crucial insights into sponsorship and review bias in e-commerce
Product reviews are increasingly important for both consumers and businesses. Authentic feedback from paying customers helps other consumers make informed decisions, enables good customer service to be recognised and helps sellers do better.
But in the e-commerce age, many sellers provide paid incentives for reviewers to influence consumer perceptions and encourage sales. While some countries have laws requiring businesses to disclose incentivised reviews, there is often little capacity for follow-up by law enforcers, and compliance is low.
Research by Lingnan University in Hong Kong is providing detailed insights into the scale of the issue, as well as the broader implications for e-marketing, consumer welfare and public policy in the platform economy.
What the research shows, explains Geng Cui, professor in Lingnan’s department of marketing and international business, is that bias in e-commerce reviews is “much bigger than most people would ever realise”. “Customers often don’t realise when reviews are inauthentic or influenced by sponsorship in some way,” he says.
The problem extends beyond written reviews, to social media and video posts by influencers, of which as few as 15 per cent disclosure sponsorship.
In March 2022, Cui and colleagues published analysing product reviews on the Amazon shopping website. Overall, compliance in disclosing incentives was found to be low, even where the law required it.
But while sponsored reviews led to lower average ratings for sellers, the reviews were deemed more trustworthy and didn’t show evidence of dampening sales. On the contrary, mandatory disclosures by platforms had a positive effect on review helpfulness and sales.
“Our findings suggest that when disclosure rules are followed, everybody benefits – the sellers, the regulators and of course the consumers,” says Cui. “Where sponsorship disclosure is not mandatory, we see problems: review authenticity declines and eventually become meaningless.”
Research shows that about 50 per cent of economic activity is based on trust – “the signals we get from each other’s experiences”, Cui explains. “It’s important we have trust. At the same time, we don't want to over-regulate or create a Big Brother society.”
But here, he believes technology will also bring solutions, for example with AI-based algorithms that can disclose inauthentic reviews to consumers.
Alongside his research, Cui teaches two courses at the university, including a cluster course on e-commerce that is open to all Lingnan students. “This is the internet generation; I learn from the students as well,” he says.
Cui is currently working with postgraduate researchers on text analysis metrics for written reviews, as well as biometric data to assess trustworthiness in video review content.
“The research is important because, as AI applications become more sophisticated, there will be bigger and more complicated questions to answer,” he says.
Going forward, he believes it is in the industry’s “best interests” that sellers and regulators work together to build a consensus on mandatory disclosure of incentivisation.
“We can hope to achieve a kind of competitive equilibrium: the honest platforms and reviewers will eventually prevail in the long run, creating a virtuous cycle,” he concludes.
about the department of marketing and international business at Lingnan University.