Raised in the business, Travis Mlakar and his father struck a deal after college. Get a job and a raise somewhere else before returning to the family business. After a few years of building a business during the dot-com boom, Mlakar moved back to Cleveland from San Diego, returning to his roots in the paper business. Mlakar is now president of Millcraft Paper, a fourth-generation family-owned business founded in 1920.  

Since taking the reins from his father in 2008, Mlakar has built a solid team of professionals, accelerated their growth platform, completed several acquisitions and implemented AI tools for order automation — increasing touchless ordering to 60%.  

Ian Heller and Jonathan Bein, Ph.D., talked with Mlakar about how artificial intelligence can make a real-world difference in mid-size distribution businesses.  

Jonathan Bein: Tell us a little about your business, what you sell and your value proposition.  

Travis Mlakar: We’re a regional distributor serving the commercial printing industry in Ohio, Michigan, Indiana, Kentucky, Tennessee and Illinois. That is about 60% of our business, the other 40% is packaging. We sell magazines, catalogs, point-of-purchase displays and direct mail. We’ve changed over time as our customers have evolved and sell packaging material, everything from corrugated to high-end boxes and box material that would go into a Tiffany’s-type jewelry box.  

Much of that is consumer-driven, high-end retail-type products. It’s been a family relationship-driven business over the last 100 years. Many of our customers are entrepreneurs and family businesses themselves. We’ve been fortunate to have customers not just for 10 years or 20 years, but for multiple generations.   

Now, we are strengthening this relationship-driven business with data and information. To do that, we’ve had to adjust our business. The mechanics of it are still the same. We do about 500,000 warehouse stock lines a year. That’s high-volume business with very little margin for error. Over time, we’ve recognized that it can’t just be about day-to-day delivery. But how can we help customers be smarter about their business.  

Our end-users are buying the substrates; they want to control what the products are. Many of them have sustainability initiatives, considering carbon footprints, so they are looking at recycled material. We help them achieve their corporate missions.  

Bein: You’ve created an interesting business from an AI tech stack standpoint. Your tech stack would be representative of a company much larger than you as a mid-size distributor. Let’s talk about how you are applying AI in your business now and perhaps also where you might be going in the future.  

Mlakar: The number one place that we have implemented AI in our business is in order entry. Much of our orders are emailed to us on a daily basis as PDF attachments. We have deployed technology that opens those PDF attachments. The AI model goes in and we can train it to understand where the PO number is, where the quantity is, where the product item number is. It extracts that information, creates the order or the data set, integrates it into our system and automatically creates the order. Our customer service reps don’t need to do any data entry.   

Bein: What are the benefits of that for you and for your customer?  

Mlakar: About 60% of our inbound orders are from electronic means. About 20% come from our ecommerce site where customers order online. We have some EDI. We have about 30% of all transactions running through the AI model I described. It frees up our customer service team. Instead of being reactive and non-value added, they can be more proactive with customers and spend their time doing outbound calling and reaching out to customers in a proactive manner, rather than just reacting and re-keying in what the customer has already sent us.   

Bein: Does this create a better customer experience and create efficiency for Millcraft?  

Mlakar: What we’ve seen from a customer experience perspective is that our mistakes are down utilizing AI. Our ability to be proactive and address problems before they become big issues has gone up. Our ability to call customers and ask for that add-on order has become a big part of our business now, and it’s proven to be very profitable.   

Bein: You said your touchless orders are at 60% now. With the cost of order entry around $8-12, if you were to evaluate the ROI, that would be huge.  

Mlakar: It is huge. And I believe there is a negative connotation with AI resulting in job loss. It doesn’t have to. We’ve found that we can take some incredibly talented people that know customers, that know products and redeploy them into other areas. We have whiteboards with all these ideas on where we can improve. Budgets don’t usually let us go after it. This opens opportunity to accelerate the change in our business.  

Bein: Were any employees threatened or upset that they didn’t have to enter orders anymore?   

Mlakar: Absolutely. They were threatened. Some people thought that that was their value and “If I’m not entering that order, how am I going to be sure that it’s handled for our customer the right way?” 

Interestingly, fast-forward nine months, and we are going through an ERP implementation, and those same people are now saying, “We’ve got to make sure this works because we have to get those orders integrated properly into the new system.”  

They recognize that they can add so much more value by proactively reaching out to the customers and predicting needs, questions and thoughts.  

Ian Heller: You’re saying their concern was that they thought customers might be let down by the new system. They weren’t so much worried about their jobs, they were worried about if the software could do the job, right?   

Mlakar: Correct.  

Bein: That is really insightful. It says that your employees care a lot about your customers. Is there another area you have implemented AI or are you considering deploying it elsewhere?  

Mlakar: We see pricing as a huge opportunity for us as we begin to segment customers and look at where there are price optimization opportunities. I think one of your sponsors, the ORS Group, looks at how you can maximize inventory and demand forecasting. I think there’s an opportunity for margin improvement and working capital reduction in our business. 

Bein: Can you elaborate on what AI is doing in terms of segmentation. Is it creating customer groups clusters?  

Mlakar: In this initial journey, we’re going to feed into the model the segments that exist and then allow the model to analyze where the price elasticity is. At what point do we begin to see a fall-off in volume? Are we seeing a fall-off in volume for specific segments or specific products? Are they being driven by market changes that we’re unaware of? Is there a competitive product in the market that we’re unaware of? We’re starting to see some leading indicators. Is there a way that the AI model can see some of those trends before we do?   

Bein: Are you doing this initiative internally or do you hire AI people or third-party developers?  

Mlakar: As part of this ERP project, we’re trying to build out the foundations, the hooks and the data elements to be able to feed into these models. Several years ago, we brought somebody on board and his job is literally innovation. A lot of his work has been on the IT side, particularly with AI. He’s hiring a team now that he will train on these tools, so that we can build that in-house. 

Bein: Several years ago; you are ahead of the curve. 

Mlakar: It can take a long time to get around the curve.  

Bein: What are your thoughts on how AI can help demand forecasting and inventory?  

Mlakar: I think that’s the answer that we’re all looking for. What is the customer going to buy tomorrow? What is that optimal inventory mix that I need to have on the floor at any given time? AI is getting to the point now where I think we’re going to be able to answer that question far better than we have in the past. You’re going to be able to look at seasonal patterns in our business. Presidential elections are a huge spike in demand just because of the amount of direct mail pieces that we all get in the mail, whether it’s state, national or local elections. Feeding that all in and trying to predict where those demand patterns are going to be; it’s not something that we’ve been very good at. The tools are there in Excel, but I don’t know if we’ve leveraged them as well as we could have in the past. I think AI is going to help accelerate that far more in the future.   

Bein: Will that mean better balance of stock and reduction in working capital? 

Mlakar: We always measure stockouts. We have a turn and earn index that we look at for most of our SKUs. Optimizing those types of metrics and particularly where interest rates are today, there’s significant money. It was one thing when money was free and interest rates were near zero, but in today’s environment, there’s real juice to be squeezed from that rock.   

Bein: You’ve set a tone that adopting technology can be used in conjunction with relationships. It’s not analytics or relationships. It’s analytic and relationships.  

Mlakar: You have to paint a picture of tomorrow and get people to understand that it can be better. We’re going to be able to use AI to ensure that we have the right products on the right floor at the right quantity for our customers so that they can improve their turn times and increase their service level with their customers. They can make bigger, bolder, better promises to help them grow their business. And all because we’re leveraging investments in a foundation so that our customers can build their future.   

Listen to our Wholesale Change conversation with Travis Mlakar, now on demand.