Geartrade, the online resale platform for used gear and apparel, is leveraging artificial intelligence not to replace its workforce, but to make its employees’ lives easier, according to company executives.
In January, Geartrade launched an AI system that was constructed with data from more than a million pieces of used gear listed, and the company claims it has seen the quality of work improve from employees as processing times have sped up.
The Daily spoke with Aaron Provine, president of the Salt Lake City-based company, about how the company designed the AI tool, how it’s saving the business money, and how it helps with current staffing challenges in the industry.
Prior to the AI tool, an employee would take each item that came in, verify it, analyze it, then build out a SKU, set the price, photograph it, and write copy for the listing.
“It was a very involved, high-touch process,” Provine said. “Our challenge was people can only get so much done in a day.”
To process the items, the worker used a combination of data entry and gear expertise.
“You have to know gear, you have to know brands and products,” Provine said. “You’re also building out SKUs, which over time just wears people down and becomes very monotonous and very difficult.”
Geartrade’s consignment intake team complained of burnout and exhaustion, all while more and more products were coming in. Plus, an employee could only process so many products per day.
Provine also noticed the productivity curve peaked in the morning then slowly declined as the day went on.
“There was also a customer impact, where they were sending us a bunch of stuff, and our queue kept getting bigger and bigger and bigger,” he said. “At one point we were nine weeks out to process an item, which was unacceptable.”
Provine’s staff was going through waves where they would work hard for a month or two then burn out on the data entry and start looking to leave.
“We had a high, high turnover rate,” Provine said. “Our employees expected more out of us. We weren’t meeting the bar for them. We had to go back to the drawing board and rethink how we approached the world.”
Provine pointed out that businesses across many sectors in the U.S. are struggling with staffing challenges.
“We’re not immune to that,” he said. “Hiring and retaining employees is a challenge. We wanted to make sure we were retaining people and keeping people engaged.”
After identifying the problem, Geartrade started looking at AI from different angles.
“People fall on multiple spectrums,” Provine said. “Either it’s going to end humanity and it’s going to make everybody unemployed or it’s going to help people.”
The tool was designed to help with two needs: Visually identifying every product that is submitted by sellers through the consignment program, and researching the item’s key features, benefits, and how much it sells for.
With the new AI-enhanced workflow, a Geartrade employee takes a photo of an item to identify the product, then the AI program builds the SKU, creates the item within the system, taps into an automated pricing algorithm that suggests a price, and uses an automated internal tool to write the copy.
Geartrade spent half a million dollars designing this program, and it’s saving them significant time and money, according to Provine.
He said prior to using this tool, building out and listing a SKU cost about $14. Now it costs his company about $4. Then there’s the impact on the employees.
“We’ve taken our ability to process SKUs from what was five SKUs an hour to 20-25 SKUs an hour per person,” he said. “Employees are much more engaged. They’re more productive. Less burnout. Overall, it’s been this tool that we’ve used to help people do their job more effectively.”
By cutting down on the time it takes to process SKUs, the company has also cut down the nine-week queue to about two and a half weeks.
According to Geartrade, since the system launched in January 2023, processing time has dropped 64%, and is projected to drop even more as the staff continues to gain proficiency with the tool.
Provine added that using AI means that the company can free up its staff up to complete tasks that were not getting completed and AI could not solve. Staffing levels haven’t changed. Essentially, the AI system is helping scale a startup business without exhausting human resources.
How They Built It
Geartrade’s internal development team built the AI program over about four months, according to Provine. The concept phase took about three months, then the development and implementation took another month.
Provine said it has been “100%” a good investment.
“We’ve solved two major issues. Employees aren’t leaving. They’re not like a hamster in a wheel,” he said. “And from a customer expectation standpoint, we’re two and a half weeks out.”
Provine expects as the program develops and they get better at using it, the savings of both time and money will continue to grow.
“Our ultimate goal is when a customer sends us an item, we want to make sure that we’re getting that item listed as fast as possible, and make sure they’re making money as fast as possible,” he said. “If we do that, we feel like we’re going to be successful.”
Bart Schaneman can be reached at firstname.lastname@example.org.