Benj Cohen knows his way around two different worlds. His family runs Benco Dental, a large dental supplies distributor in Pittston, PA, and from a young age he spent plenty of time working in the warehouse and office, eventually going out on sales calls with his father and the other outside reps.
At Harvard University, Cohen blended his experience at Benco Dental with a degree in applied math to found proton.ai (www.proton.ai), a company dedicated to bringing artificial intelligence to the B2B world to deliver large ROI.
While working on some data analytics projects during his junior year of high school, he started to see potential applications for artificial intelligence in the distribution industry. It wasn’t until a year ago, in his junior year in college, with the help of a small but talented team of developers, Cohen launched proton.ai. He likes to joke that the company’s data scientists are way more technically capable than him.
We met Benj at the National Association of Wholesaler-Distributors (NAW) Annual Conference earlier this year in Washington, D.C., and it was obvious from our first conversation with him that he had a down-to-earth sense of how to harness technology to increase distributors’ sales, one customer touch at a time.
Cohen’s take is in dramatic contrast to some technology companies in the dot-com era that came into the distribution world without any related industry experience, but with the intention to tear the distribution business down and rewire it with their own technologies. To that he says, “If only it were that simple. Instead we obsess on solving the challenges our customers face.”
Cohen says the wholesale distribution business has several unique dynamics that his team had to consider when designing their platform — the massive amount of data that distributors must deal with; the huge numbers of products they typically stock (his family’s business, Benco Dental stocks about 200,000 SKUs); and a degree of scale to manage it all.
The key, he says is helping distributors mix and match the products to customer needs. “In the distribution business, there are more combinations of customers and products than there are stars in the universe. By a lot, actually,” he says. “What that means is it’s hard to find the right thing for every single customer. We believe this type of personalization is just the beginning of AI’s impact in wholesaling.”
But it’s not as simple as just pairing customers with products. “It’s personalizing every customer touch using AI,” he says. “Part of that is finding the right product and the right customer, but it’s a lot more than that. It’s figuring out which customers to talk to when, what to talk about to each customer, and how to talk to each, over which channels. That’s where we see artificial intelligence having the biggest impact with wholesalers.”
Some of the key AI capabilities that proton.ai offers include:
- If a customer buys a product every month, proton.ai’s Cyclic Purchases capabilities recognizes if a customer buys a product every month and will recommend the product at reorder time.
- The company’s AI capabilities find items the customer is likely to purchase but hasn’t yet. proton.ai says these recommendations are capable of increasing revenue per customer touch by nine times.
- proton.ai builds a database of similar products, facilitating upselling, cross-selling and product substitutions
Cohen believes the major players in the business-to-consumer world (B2C), like Amazon and Netflix, have got it pretty much right as far as customer personalization, and that the B2B world can learn from the B2C model.
His team at proton.ai hopes to blend their experience developing search algorithms for business-to-consumer websites with a knowledge of distribution. “We’re leveraging our team’s experience building patented personalization technology for Fortune 100 B2C players along with experience working at Amazon and IBM to bring deep personalization to wholesale,” Cohen says.
“B2C personalization, however, does not work out of the box for B2B. There are more re-orders, the purchases are more complicated and the products are more technical. There are a lot of pieces and reasons why out-of-the-box B2C solutions don’t work right away. This creates a moat around wholesaling.”
According to Cohen, one of the biggest pieces missing from the business-to-business world is product suggestions. “Customers understand the personalization that Amazon delivers,” he says. “We deliver Amazon-quality artificial intelligence to B2B distributors go-to-market channels. Amazon makes 35% of it’s e-commerce revenue from AI powered product suggestions. People may think about product suggestions and upsells as a nice-to-have, but for companies like Amazon they create billions in revenue and a powerfully differentiated customer experience.”
It may seem that artificial intelligence is collecting so much information on products and customers that salespeople won’t be as necessary in the future, but Cohen says this is absolutely not true. He believes salespeople in the distribution industry will always have a role because they have relationships and knowledge that can’t easily be modeled by an AI system. “They are making that final-mile decision,” he says. “Artificial intelligence will not replace reps, but rather supercharge them to deliver an even better experience to their customers.”
Cohen envisions that proton.ai’s platform will also offer supply chain and inventory management capabilities that help distributors better forecast exactly what customers will buy. “We imagine that in the future, our system will tell salespeople what to talk about, customize content online for customers, and manage inventory, stocking products for a specific customer before they even buy it,” he says
Cohen also sees artificial intelligence impacting other business functions at distributorships, such as streamlining the hiring process by selecting resumes that stand out, although he feels AI’s impact will be much greater in places where the technology touches customers.
Cohen says the results of proton.ai’s early installations are promising and that in one trial they were able to dramatically increase the amount of revenue sales reps were generating per customer touch. “More people should be paying attention to the revenue per customer touch metric because it’s the driving force of incremental growth,” he says. “Distributors need to focus on how much revenue they make per customer touch and how they can increase it.”
“You only have a certain number of customer touches you can deliver, so you want to figure out how to increase the revenue generated by each one.” Cohen believes artificial intelligence can deliver the automated personalization to increase revenue per touch, turbo-charging distributors.