Nine years ago, as a nursing student, Curtis Boyd agreed to help a doctor fight a slew of fake negative reviews about his practice. He never imagined that single task would shift his career trajectory entirely.
Boyd decided to go back to school to study coding, machine learning and artificial intelligence and eventually created a platform to help open consumers' eyes to fabricated online reviews.
The Transparency Co. launched in September and is designed to scan review sites like Google and Yelp and collect raw data that is analyzed. The platform then makes predictions about whether a business's positive reviews are real or fabricated using machine learning and artificial intelligence.
"We’re living in a time where reviews are so accessible," Boyd said. "You can go online, spend 100 bucks and have the same reputation as someone who’s been in business for years, and it’s really hard for consumers to see who’s the better choice because these (companies) have the same number of reviews and the same star rating."
The platform operates through a free Google Chrome extension that is compatible with any Google domain. Boyd said hundreds of individual consumers, business owners and marketing agencies have used The Transparency Co.'s services since its launch. The startup recently inked a government contract in Arizona, though Boyd was unable to disclose details of the deal.
He said the data is labeled on a transparency report that contains the evidence to show with exceptional accuracy whether a business's reviews are real or fake. The data page sorts the reviews to let the person reading the report know which are real and which are not. It also points out clusters of reviews, which are reviews written by one person and typically an indicator of fabrication.
"Usually in this situation, you have one person using multiple profiles to create more content," he said. "A large number of anonymous reviews are also a red flag. Typically, only 2% of real reviewers do so anonymously."
Additionally, a business's data report includes a map that shows where the reviews came from. Multiple reviews from all over the globe is another major indicator of fabrication, Boyd said.
"If this was Trip Advisor, this would be no big deal because it’s normal for people (from all over) to write reviews when they travel," he said. "On Google for something like a home service company, it should not look like this."
