Fake or inauthentic online reviews have been a problem for years, but artificial intelligence is making the problem even more complicated. Many customers are wondering if reviews are still reliable. Many businesses are wondering if reviews are still valuable. AI is undoubtedly changing the landscape of online reviews. But what can we expect? And how do we know?
Is AI Killing Online Reviews?
AI has already changed online reviewing, and it’s likely to continue changing. Looking at the state of AI and reviews now, the tools available, and how these tools are improving, we can make some predictions.
The Problem With Fake Reviews
Between 80 and 90% of customers rely on reviews before making a purchase or visiting a business. Online reviews have become vital waypoints showing customers which businesses, products, or services are right for them, what problems they might encounter, or what expectations are realistic.
Of course, the importance of reviews gives businesses incentives to fabricate them, or pay others to create them. The use of fake and paid reviews has grown quickly as more and more consumers look for products and services online. This practice eats away at the reliability and authenticity of online reviews overall. If customers feel that they can’t trust or believe online reviews, their value falls.
Understanding the importance that genuine customer reviews play in their business model, Amazon recently sued over 10,000 groups providing review-for-purchase services. This isn’t the first time in recent memory Amazon has sued large groups of paid reviewers, and the number keeps getting larger with each suit. Other businesses that rely on honest reviews, like Yelp and Google, have filed similar actions, and updated their terms of service years ago to strictly forbid incentivizing reviews. Still, the practice persists.
Video reviews are more authentic and much harder to fake than text alone
Boast makes collecting video reviews easy >
AI Supercharges Fake Reviews
Whether they’re posted on Amazon, Google, Yelp, or another site, reviews generally aren’t very long. This means it doesn’t take long for fake reviews to make something up. However, artificial intelligence programs give bad actors a chance to supercharge their efforts. AI can generate fake reviews in seconds, which even the fastest paid reviewers can’t quite manage. AI programs even remove the need for additional reviewers; businesses selling reviews can simply create fake accounts and farm the actual review-writing out to AI.
All this makes it easier than ever to create disingenuous reviews. AI is already clearly generating reviews on sites like Amazon. Some reviews even begin with Chat GPT’s well-known phrase, “Yes, as an A.I. language model, I can…”
However, this doesn’t mean AI will destroy online reviews. The solution, surprisingly, may also be AI-generated.
AI Detects Fake Reviews, Too
Unfortunately, humans have a hard time telling which reviews are genuine, and made by real people, and which are fakes made AI programs. Studies show that humans only guess right about half the time, indicating our intuition isn’t much better than chance. However, AI programs are experts at pattern recognition and data analysis. The same mechanisms that allow AI programs to imitate human writing also allows them to recognize other AI-generated work.
Several AI tools already exist for uncovering AI-generated reviews. These tools use a few different mechanisms to recognize their AI-counterparts. They use language patterns, the information contained in the review, connections between reviews and reviewers, and behaviors, like when a review is posted, to root out false reviews. One of these indicators by itself might not be reliable, but the AI programs weigh each of these indicators, and use vast amounts of data to recognize patterns of authenticity and falsehood.
Though our own ability to recognize fake reviews isn’t great, nor is it quick, AI programs perform better, and they can root out fake reviews much faster. One study found that an AI program detected spam reviews with about 96% accuracy. This means the businesses that host reviews, like Amazon, as well as organizations in charge of advertising authenticity, like the FTC, will have powerful tools to fight back against fake AI reviews.
AI Offers Other Benefits
Though AI has certainly made the fake review problem more complex, these programs may also offer some other capabilities. Some AI tools are using customer reviews to better understand customer sentiment, shifts in customer attitudes, trends, and even identify opportunities for better products and services. These customer analytics AI tools allow businesses to draw insights from customers faster and more accurately. Instead of costly surveys or market research reports, businesses may soon be able to draw meaningful insights from consumer commentary on review sites, social media, and more.
How to Improve Authenticity in Reviews
Businesses may be understandably concerned about whether or not their own reviews are believable. Many reviews are short and may sound generic, even when they’re made by a real customer. However, other customers may find these reviews suspicious. With too many of these, customers may find the business in general to be suspect.
There are a few ways that businesses can combat the growing surge of fake and online-generated reviews. Encouraging customers to leave reviews that sound more authentic can help show other customers that your business truly delivers on its promises. Here’s a few strategies you can use to get more authentic, human-sounding reviews:
- Use video: Though AI programs can create videos, they’re far from convincing yet. Video reviews or testimonials from real customers are much more authentic and effective.
- Provide prompts: If customers don’t know what to say in their review, they’re likely to provide a generic response. Provide a few prompts like, “tell us about your favorite menu item,” or “tell us about why you visited us today.”
- Forge connections: Customers with deeper connections to your brand are more likely to leave detailed, positive comments, reviews, or testimonials. They’re also more likely to purchase again, tell their friends, and stay with your business in the future. Building a loyal customer base takes time and money, but it can be worth it.
- Use the tools available: Obviously, it’s a bad idea to create fake reviews with AI programs. However, you can use automation tools to ask real customers for reviews quickly and efficiently. You don’t have to break the rules or do something unethical to make the review-gathering process a little easier.
AI may change the online reviewing landscape, but it won’t destroy it. Online reviews will continue to be an important part of customers’ searching and decision-making process. Businesses that continue to provide positive customer experiences and ask their customers for reviews will continue to see value from them.
Posted in: Online Reviews