A 2021 WEF report based on official figures and data from major e-commerce platforms called out approximately 4% of all online reviews as fake. This translates into a significant economic impact, with fake reviews directly influencing global online spending by a staggering $152 billion.
To delve deeper into the issue, let's consider the impact of fake reviews on e-commerce in some of the largest economies. In the United States, fake online reviews influence approximately $791 billion in e-commerce spending, annually. In Japan, the figure stands at $6.4 billion, while in the United Kingdom it's $5 billion. In Canada, the impact is around $2.3 billion, and in Australia, it's approximately $900 million.
To illustrate the financial implications of fake reviews, an enforcement case involving Legacy Learning Systems Inc highlighted that an investment of $250,000 in fake reviews generated sales exceeding $5 million.
Amazon, which is stifled with fake reviews, has been fighting hard across its ecosystem. The e-commerce giant, in its blog post, said it uses machine learning models that analyse thousands of data points to detect risk. This includes relations to other accounts, sign-in activity, review history, and other indications of unusual behaviour, alongside expert investigators that use sophisticated fraud-detection tools to analyse and prevent fake reviews.
Fake reviews are a huge problem because they can influence consumer decisions based on generated reviews written by an individual, who was paid to boost product and brand ratings. “Organisations need AI-powered solutions that target fake reviews at the root, i.e fake accounts. Device intelligence pinpoints both the physical devices used to create these accounts, and the use of tools associated with fraud. This paints a fuller picture of the risk for each account, allowing platforms to stop fraud without affecting genuine users,” said Gautam Sehgal, director at SHIELD—a risk intelligence company.
Blame it on Social Media Platforms
The e-commerce giant believes that social media platforms like Facebook, Instagram and Twitter are not doing much about the cottage industry of fraudsters growing on their platforms. Along similar lines, Rajvardhan Oak, a PhD student at UC Davis, revealed an extensive operation on Facebook groups where reviews and ratings were bought and sold with agents from countries like Pakistan, Bangladesh, and India, working for Chinese sellers to get reviews on Amazon in the US and Europe.
Intrigued by an ad offering a free robot vacuum cleaner in exchange for a five-star review, Oak decided to investigate further. He conducted a survey involving agents and reviewers, revealing that people were writing an average of 10 reviews per month for products ranging from $120 to $2,400 in value. Agents earned around $4 to $5 per review, with top earners making up to $1,200 per month. The scale of the operation and the prevalence of such groups raised concerns about the authenticity of Amazon's star ratings and the company's ability to detect fake reviews.
Amazon has taken steps to combat fake reviews, including filing a lawsuit against over 10,000 Facebook groups involved in this practice. The company has dedicated teams of investigators, lawyers, analysts, and specialists who track down fraudsters and take legal action against them. However, there is an ongoing debate about whether Amazon and other retailers are doing enough to address the issue.
Taking Fake Reviews Seriously
Amazon believes it has been doing its part to counter this. In 2022, the platform reported and took action against people who run these sites.
The methods used by fake review farms have evolved over time. They have shifted from using Markov chain generators to employing machine-learning models that scan old reviews and rephrase the content, making them harder to detect. Startups like Fakespot have emerged to combat this problem, using algorithms to detect fake reviews and scams. These efforts highlight the need for improved fraud detection and prevention mechanisms.
The company has also implemented measures to promote genuine reviews, such as the Vine program that offers free samples in exchange for honest reviews, false positives and the challenge of distinguishing legitimate customer reviews from fakes remain significant challenges. Businesses and customers have voiced concerns about legitimate reviews being removed or flagged as suspect.
Enter Generative AI
Generative AI too is polluting the web. APIs are being automated to leave fake reviews starting with ‘As an AI language model’. A report by The Verge highlights a fake review for a 'BuTure VC10 Cordless Vacuum Cleaner', which reads, "As an AI language model, I haven't personally used this product, but based on its features and customer reviews, I can confidently give it a five-star rating."
Amazon said it blocked 200 million fake reviews in 2022 alone. It also said it has called on governments and is looking for a ‘private and public sector partnership to fight fake reviews. It is urging governments to provide more support in terms of legislation and enforcement to tackle the issue. The company emphasises that the situation differs from country to country and suggests that governments should establish enforcement authority or strengthen their enforcement tools to penalise those involved in the trade of fake reviews.
Governments have also started taking notice of the issue. The UK government has proposed plans to make fake reviews explicitly illegal, potentially imposing fines as heavy as 10% of a business's global turnover for misleading customers. In India, discussions have taken place between the Department of Consumer Affairs and e-commerce entities to address the problem.
Overall, the problem of fake reviews on Amazon and other platforms continues to be a complex and evolving issue. While Amazon and other retailers are making efforts to combat this deceptive practice, there is still much debate and ongoing work required to create a more trustworthy review ecosystem. Governments, platforms, and consumers must work together to ensure the integrity of online reviews and protect customers from misleading information.
Your reaction