While the existence of fake reviews on e-commerce sites is well-known, the extent of their prevalence as well as the impact they have on consumers’ purchasing decisions is less certain. As such, in a project commissioned by the Department for Business and Trade, Alma Economics sought to shed light on this issue.

To determine the prevalence of fake reviews, we built a machine learning model that was trained on a dataset of known fake reviews. By subsequently applying the trained model on a dataset of 2.1 million product reviews across 9 popular UK e-commerce platforms, we were able to estimate the percentage of product reviews on these platforms that were fake. We found that approximately 11% to 14% of all reviews for common products are fake.

We then investigated the effect that these reviews have on consumers when they make purchases online. To do this, we had nearly 5000 participants take part in an online shopping task, an experiment that was built by our talented developer team. The participants, all familiar with shopping online, were asked to purchase one item (out of three) on the platform based on the information that we presented to them. They were unaware that we varied this information between groups of participants. Specifically we:

Embedded poorly written or well-written fake reviews among genuine product reviews for one of the three products displayed for some participants
Included a “warning” textbox that steps had been taken to moderate misleading content (including reviews) on the platform for some participants.

By doing this we tested whether the different types of fake reviews and/or the warning textbox impacted the choices that consumers made online. Our key findings were that:

Poorly written fake reviews decreased the likelihood that a product would be purchased by 5.3%.
Well-written fake reviews increased the likelihood that a product would be purchased by 3.1%.
Informing consumers that steps were taken to moderate misleading content on a platform does not impact consumer purchasing behaviour.

We also conclude that fake reviews cause an estimated £50 million to £312 million in total harm to UK consumers. However, this estimate does not include the impact of fake reviews on consumers who purchase services, on future consumer behaviour or the separate impact of inflated star ratings (which often accompany fake reviews). As a result, this is a conservative estimate and the true consumer detriment arising from fake reviews is likely to be higher.

Our findings suggest that consumers are more susceptible to being misled by well-written fake reviews than poorly written fake reviews. If consumers are generally not able to distinguish between genuine and well-written fake reviews, and fake reviews are becoming more sophisticated and difficult to detect over time, the negative impacts of fake reviews on consumers are likely to increase over time. Furthermore, the finding that participants are not affected by a warning about misleading content on the platform suggests that consumer trust in product reviews may be difficult to alter.

To read and download the full report and executive summary click here.

The impact of online fake reviews