Bogus reviews on Amazon, Google, websites, and other review hubs are unfortunately not a new phenomenon and were a problem for businesses and their customers long before ChatGPT was even a twinkle in a programmer’s eye. However, on Amazon over the past year, the number of fake AI-generated reviews has exploded by up to 400% compared to before the launch of ChatGPT in November 2022. As one of the world’s most respected customer review sites, this is a huge problem for Amazon and the B2B and B2C businesses that use the platform.
In August this year, Amazon announced the launch of a new customer review scanning service (also using an AI algorithm) to identify and remove false reviews created by generative AI, but even so, an estimated 8% of all Amazon reviews created in 2023 are now AI-generated, up from 2% in 2022.
What About Reviews On Google, Tripadvisor, And Other Sites?
Unfortunately, fake AI reviews are a problem on these sources too, and despite a stern position being taken by Google and others against the practice, it’s a difficult one to police. In July 2023, a Guardian journalist successfully submitted a credible-sounding but bogus AI-generated review to TripAdvisor relating to a hotel in Poland, which was accepted and published by the platform.
Firstly, every business should be concerned about the rise in AI-generated reviews, and no one should assume that it is simply the B2C or online retail sectors that are affected. Everyone stands to lose from this should the practice become widespread and unchallenged. The reason for this is straightforward. Digital marketing is based on credibility, which is increasingly underscored through social proof – independent reviews, testimonials, and case studies posted on Google, Facebook, LinkedIn, and other sources.
As more and more sales decisions are made exclusively online, prospects increasingly seek out these reviews at the awareness stage, before finding anything else out about the business, and not solely at the decision stage before a final purchase is made. This has been an extremely important change and is comparable to an employer asking for references before accepting a job application.
False reviews, whether created by software or humans, destroy the credibility of individual businesses and of the social proof system in general, making it harder for smaller businesses and start-ups without established brand authority to make sales online. It only takes one fake review to completely discredit your business, even if you have hundreds of genuine reviews published online.
So, our first point is, don’t publish fake reviews. It takes seconds to create a review on ChatGPT but the damage to your business could last for years.
But even more importantly, how do you spot an AI-generated review and how do you give your customers confidence that your reviews and social proof are genuine?
AI reviews are simply created by software, so are more likely to include unverified sources. If you use Amazon to make sales, encourage reviewers to verify their purchase, which allows the platform to check whether the customer has genuinely bought or used the item on Amazon, after which the review is labelled as a verified purchase. Google also offers a verified review service for local searches and service ads, although unfortunately, this is only currently available in the USA and Canada – we can only hope it will be expanded globally in the near future.
A 2023 survey of 26,000 Amazon product reviews found that ‘extreme’ reviews (i.e. One-star or Five-star) are 1.3 times more likely to be AI-created. This brings us back to our point that all (genuine) online reviews are useful. No matter how good your business is at customer service, you’re likely to accrue a broad range of star ratings. To give the most authentic and genuine social picture of your business and its service delivery, publish all your reviews, even the two and three-star ‘meh’ reviews that many businesses discount because it lowers their overall average star rating. A balanced review profile could, in future, give prospects greater confidence in the legitimacy of your reviews and the credibility of your business than straight fives.
To encourage genuine and non-AI generated reviews from your legitimate customers, ask them not to use ChatGPT, and to make it quick and easy for them to do you this favour without having to write an essay, consider sending them a review template or set of questions to use.
To give your prospects and customers confidence that your reviews come from legitimate sources, request that all customers who leave reviews or case studies provide traceable contact details – e.g., the name of the reviewer with a hyperlink to their LinkedIn profile – and a link to their business website which shows the company name, registered address, and limited company registration number. Try to minimise the use of anonymous case studies and reviews on your site, as these are more likely to be flagged up as potentially AI-generated.
At JDR Group, we encourage the use of social proof in all forms of digital content in order to increase the credibility and sales potential of online businesses. To find out more, please get in touch with one of our inbound marketing specialists today.
Image Source: Canva