How an AI-Powered App Helps Thrift Stores Maximize Value and Deepen Community Impact

After our reverse logistics startup was acquired, my cofounder, Josh Packard, and I were looking for what’s next. We didn't feel like we’d fit in working at a big company and a college friend and VC said that his mom’s thrift store was crushing it, making hundreds of thousands for her community each year using no tech. We looked into the market and found that thrift is a $20 billion a year industry, and with smarter pricing, it could be $30 billion.

With our backgrounds in reverse logistics, we saw how thrift was this underserved market that could really benefit from tech. The biggest need of all: figuring out the right price in the back room. Pricing decisions were being made by 10 different people per store across 25,000 stores; that’s a quarter-million people pricing stuff, with no software helping them with price discovery, ensuring consistency, and making sure the right sales channels were selected. With the advent of AI, we saw a big opportunity for a simpler, smarter way of pricing that’s informed by a mass of data. So we launched Thriftly, an app for thrift store operators that streamlines pricing, tagging, and listing items online.

Josh and I are mission-driven entrepreneurs, and that’s why this idea really stuck and became our passion. Most thrift stores have an employment mission–it's really the jobs that it creates and giving people an opportunity when others wouldn’t. What we really unlock is enabling these thrift stores to capture more of the value that they're creating by getting smarter with pricing with help from AI. Not to mention the whole sustainability aspect. We’re really motivated to find ways that we can be less wasteful, which Thriftly helps solve, too, by keeping more goods in circulation.

About 10% of a thrift store’s revenue typically comes from selling stuff online, and there are usually several people involved with researching what the thing is and its value. We saw an opportunity to make that process wicked fast using Google Lens and the Gemini model while empowering employees to price 20-50% more items, and with that satisfaction of driving even more revenue to support thrift nonprofits’ vital missions.

As we continued to grow we realized that we can help with the other $18 billion of thrift— in-store Every second counts with pricing items that sell for $5 on average. We chose Gemini after researching many different models because it’s so fast. Given the response time, we were able to unlock a whole new way of brand-based pricing for physical retail thrift stores, which is really a transformative thing.

People are still critical in the process with steps like evaluating the quality of the item, but we're shifting as much of the time-intensive work off the thrift operator’s shoulders and onto AI so they can focus on getting more items on the floor.

As we mass a large corpus of data about the items that are priced and sold, we're going to be able to start taking further steps from brand-based pricing to doing micro appraisals. I’m looking to use Google AI to help with this, too. Getting that granular about selecting the right price will be a huge unlock for thrift, making sure they’re capturing more of those funds that go into the employment and other programs they run in their communities.

On the sustainability front, we know that about half the items that go on to thrift store shelves right now get compacted and sent overseas. What if we could keep 60, 70, 80% of the items in circulation? With AI like Gemini enabling a simpler, smarter way of pricing, we can really start making a dent in overall waste.