Proton founder and CEO Benj Cohen recently joined Jason Bader on the Distribution Talk podcast to talk about how artificial intelligence is revolutionizing distribution and how distributors can take advantage of this technology.
AI requires a great deal of complex data to learn and make predictions. Cohen says that distributors are the perfect use case for AI because they have years of data for models to analyze. Here are just a few highlights from the conversation.
Bader: Most people catering to the distribution market don’t have the deep industry background that you do. Would you mind sharing a bit about your background in distribution?
Cohen: I grew up in my family’s dental distribution business. We sell products to dentists all across the U.S. My great-grandfather started the business in 1940 out of a suitcase. My grandfather went into the business, and he and his two kids, my dad and uncle, grew the business to become a national player. They’re the third-largest dental distributor in the U.S. and the largest family-owned in the market.
Bader: So, how did you get into it?
Cohen: One summer when I was in high school, I partnered with someone in the analytics department at my family business, and we worked on some basic modeling. This turned me on to data science and to the idea of leveraging data to achieve a business outcome, specifically in distribution. When I went to college and studied applied math and data science, I remembered a challenge I had seen when I traveled with my dad to dental offices. When you sell a lot of products, like all distribution businesses, it’s hard to figure out what your customers want.
The median Proton customer sells 687,000 SKUs. It’s really hard for salespeople to figure out what items their customers want to buy. The combination of using some of the cutting-edge techniques I was learning about in school, applying that to the large data set we already have, and then giving the output to salespeople. It’s the insight that led me to start building software for the inside sales team of my family’s business. We ended up increasing their close rates with the initial version of the software by four or five times.
Bader: That’s when you figured out you were on to something?
Cohen: We thought, oh man, we are onto something. If this simple thing we built could have this kind of impact, this is a real problem. We have something that can solve the problem. And it’s not just dental distributors; all distributors have this problem. That’s really where Proton began.
Bader: How receptive were the salespeople to this kind of approach?
Cohen: We focus on building a partnership between our technology and the salesperson, not replacing what the salesperson does. With that approach, we found that salespeople are really excited about using our technology, especially expert reps who see it as a way to save time, be more efficient and sell more.
Bader: Could you clarify what Proton’s approach is to AI in distribution?
Cohen: What we do is we ingest all the data. This includes the distributor’s customer, product, transaction, quote, online, browse data — truly everything we can get about what our distributors’ customers are doing. Then our AI models churn through that data to understand patterns across customers, products, and reps. They’re deep-learning models, finding extremely sophisticated patterns. And then, we use the output of that to make the rep smarter about what the customer needs to buy and make the customer smarter about what they’re looking for.
Bader: Is your product better suited to an inside sales team or to an outside sales team, or does it really matter?
Cohen: Distributors are fundamentally cross-channel businesses and sell through sales channels, including outside sales, insides sales, customer service, and e-commerce marketing. The way we think about it at Proton is to build the best possible solution for distributors; it has to be an omnichannel solution. We’ve got to make recommendations no matter where the customer is interacting with the distributor. If we think about it from an omnichannel perspective, how do we use AI to connect all of our channels together and make them all smarter about what the customer needs? So what that looks like for insides sales is different than what it looks like for outsides sales, but it has to be everywhere for it to be successful.
Bader: Is there a certain size organization that tends to adopt AI easier?
Cohen: We work with distributors anywhere from $50M to billions and billions in sales. We’ve seen success at the smaller end and the larger end. The thing to remember is distributors already have a ton of data. That makes adopting technology like this pretty easy once you have the model set up. It’s really easy to get started.
Bader: Do you need a lot of data scientists to make this work?
Cohen: I felt that if my family business didn’t have the tech that we were building to know what to sell customers, then Amazon would solve the problem better at some point. The reason I started this business was to make it easy for my family business, but also distribution in general, to be able to leverage the same kinds of technologies Amazon uses without having to have a data science team. The idea is to make it as easy as possible for distributors to get started. So, no, you don’t need a team of data scientists.
Bader: Over time, I’ve seen these advanced technologies start with bigger distributors but then find their way to smaller distributors. Is that a direction you see Proton going?
Cohen: In the long run, we’ll definitely be able to serve smaller and smaller distributors as the AI technology gets more mature and less expensive. But in terms of having enough data, smaller distributors certainly have enough. That’s really the key thing about what makes our business work. We’re starting in a market where every distributor has the data that’s required to make a system like this run.
Hear Bader and Cohen’s full chat on how AI integration can help distributors bolster sales and get ahead of the game: https://podcasts.apple.com/us/podcast/benj-cohen-of-proton-ai-on-leveraging-sales-data-with/id1465977571?i=1000541317911