Mathew Witte Interview: Dynamic Route Optimization
I recently had the pleasure of sitting down with Mat Witte from ORTEC for an in-depth discussion on route optimization best practices.
Executives still get so caught up in thoughts like if I've got 10% growth next year, I need 10% more trucks. They miss the power of planning and execution.
Mathew Witte is Senior Vice President, ORTEC Americas. In this role, Mat is responsible for ensuring ORTEC’s customers continue to grow and succeed through data-driven continuous improvement. He is also driving ORTEC’s efforts into new markets and industries across North and South America.
Mat comes to ORTEC from Mohawk Industries, where he most recently served as the vice president of Warehouse and Delivery Operations for the world’s largest flooring manufacturer. In this role, Mat was the executive in charge of all final mile warehouse and delivery operations for North America.
Prior to joining Mohawk, Mat led the consulting division of UPS Logistics Technologies, also known as Roadnet. In his time there, Mat developed his reputation as a final mile expert and developed many value-added consulting initiatives that helped define the final mile software industry.
He has an MBA in Supply Chain and Acquisitions from the University of Maryland and a BS in Management. He and his family live in Kennesaw, GA.
Bill Mathews: How would you explain static and dynamic route optimization to a novice?
Matt Witte: Yeah, it's funny, way back in the 90s, I was the much younger, practical voice of dynamic routing: back in the first-generation software days. It was 1998 when we rebuilt the algorithms, and I went out and did a lot of publicizing for that, the real world of dynamic routing, and what I found after coming back to this side with ORTEC is, it hasn't changed. Executives still get so caught up in thoughts like if I've got 10% growth next year, I need 10% more trucks. They miss the power of planning and execution. So delving into your question, we used to have it all black and white. If you’re static, you have the same stops, same places, same locations, same drivers, same orders, and the same frequency, no matter what. It's the same route every single week, or every single day. The flip side is you're fully dynamic. You throw orders into a hat and tighten them as much as humanly and mathematically possible.
I used to run executive training with 200 people in the room, and I’d say, which is more efficient, Static or Dynamic? And whatever they would say, the reality is they’re both wrong. Because either end of that spectrum is incredibly inefficient. We're all in a dynamic world. Every customer does not order the same volume every week. Every customer doesn't have the same needs every week. You don't have the same tractor or trailer available every single day. You don't have the same driver skill, so that pulls you off that left, static end whether you like it or not. For the pure dynamic side, the fact that you would ignore time Windows, truck types, and keeping drivers in the same area, nobody wants that either.
So for most industries, they need to be somewhere in the middle. Where do you end up on that spectrum? It depends on the restrictions that you have. And, as an organization, you must understand the constraints that are important for your business. Am I giving my top customers two-hour time windows? Am I getting four-hour time windows? Is everybody getting day-long windows? Are there key accounts? Is there a driver skill issue? Is there a tractor requirement? Do some customers have to be the first or last stop? There could be accounting or credit-type issues or frequency and service level. As you review all these, they will tell you where you end up on that spectrum of static versus dynamic. Hopefully, you're leaning pretty far towards dynamic, within reason.
Bill Mathews: What about the concern of driver familiarity with the territory? I think it can be a little overblown if you're typically going to go into those territories every day regardless, and even though the stops and the stop orders may be dynamic, you can still put a driver into a familiar area. You can even constrain dynamic routes by area.
Mat Witte: Sure, that's what I found. I’ve got a unique background. I started my career in the beer industry supply chain. Then I went into a consulting role with a first-generation routing provider, and the biggest reason I left that consulting world and went back into the supply chain, is that everything I was teaching was getting great results with Fortune 100 companies. I took all that I was teaching and applied it myself at one of the world’s largest manufacturers. I found that you're exactly right. Even in a perfectly standard world with the same driver, the same stops, drivers go on vacation, and drivers get sick. You know, there's just too much volume some days or there's no volume – you're going to change. So what I found when I was going to almost pure dynamic, you know as a supply chain guy in a Fortune 300 company, was that I would keep a primary driver and a secondary driver in an area. And the beauty of the customer service side of that, especially in specific industries like food service or beverage, is that the driver relationship matters. They can get in and out of a stop much quicker than a new driver while providing better service, and it speeds up the service. It gives the customer that familiarity, but again, you don't do it with just one driver. You're always going to have two, sometimes three.
Bill Mathews: So, who would you define as the ideal customer for a dynamic approach?
Matt Witte: You know, it depends on the technology we're using. The higher the density, the more fun the Dynamic approach will be and with better savings. You're seeing beverage companies that have 30 and 40 stops within a few miles radius. That's where pure dynamic works best. Anybody that has to manage service frequencies, which is a whole different potential white paper for you, if you have frequency optimization or you are managing the service days to better enable your dynamic routing, that works too. But for anybody that has a higher density they can be more dynamic and achieve more savings.
Many people may think the opposite for routes that don’t have customer contact. If you're doing doorstep deliveries where you don't talk to people, they would think that they need to be dynamic and there's a point there, but again, you still have familiarity issues. But for me, I like higher density.
Bill Mathews: One of the things that I always encountered in a dynamic routing environment is this: the planners are bringing up these routes and they're looking at them all on the map. On a dynamic, time-window-oriented day, those routes could look very much like a plate of spaghetti. The John Henrys, as I refer to them, of the route planning world, are trying to beat the steam hammer. They're gonna go in there and spend the next three hours undoing all those crossing points and looking for a pretty flower petal. I don't know how much you've encountered that. Because to me the human element always comes into play hard on both sides of this spectrum. Because if they're fully static, then they're spending the next two to three hours unloading that static plan into something that will work for the day.
Mat Witte: Yeah, that's nearly universal, and I see it all the time. I gave a speech about two weeks ago using this slide from the folks at Boston Consulting, the source of the slide was American Quality Center or something, and they talked about On-Time-In-Full (OTIF), across the Fortune 500 from 2010 to January 2020. Before Covid.
This slide and the supporting paper show that OTIF is lower now than it was back in 2010.
However, inventory is higher, and headcount is higher! So they're putting more stock, they're bringing more people, yet. They're performing worse. It was very descriptive of exactly what you're asking in that if I go back 15 years ago or even 10 years ago, what we talked about in the consulting world was data: we need more data. If we have more data, we can make better decisions. Now, guess what? We have data. There's so much crap out there that I think people are frozen.
I very clearly see that people that have first-generation software packages out there, they've been sitting out there for something like 20 to 30 years, that they have lost adoption. They've lost data credibility and they take this tool, and I see it all the time, they manually move stops around to accommodate for bad data. So they take a dynamic answer and they make it more of a standard because they don't trust their data, they're just talking to a random person at night. Then the turnover in that routing profession is very high. These people haven't been professionally trained for two or three generations. That's why companies lose adoption, and it is across the board. I had a meeting with a Fortune 50 company just a few weeks ago. I watched these routers and saw the same exact thing I've seen all over and I know you're seeing it.
BIll Mathews: I've seen it over my 25 years, probably 98 to 99% of routing systems are used as a glorified whiteboard. You can make a ton of arguments why that is the case, but certainly data's primary. And, the human element I think is number two. Where they're looking at this and the drivers complain that they saw another truck out on their route for example. No one understands the whole picture enough, none of it makes sense to them. They're not getting to the bigger picture. That's why you're not as efficient. Training and change management are missing from these scenarios
Mat Witte: And I see it's getting so bad that people at the executive level don't understand it. Some “optimization” companies out there are doing no optimization at all. They've dumbed it down and they are successful by just doing a glorified zone and sequence with a fancy user interface. They are replacing legacy systems all the time.
Now you've jumped the shark and gone backward. Companies are actually making investments to go backward. But, I also blame the software companies as well because all these software companies over the years have sold the product and left the customer on their own for training. That's where you and I come from. We want to change that, knowing that the special companies are the ones that take advantage of it.
Bill Mathews: Last question then, what do you see as the true advantage of dynamic optimization?
Mat Witte: It is at an executive level, and I talk to CEOs all the time. This optimization will allow you to grow and scale better, absorb costs, and absorb capacity. There's this trifecta of efficiency, And it's the same, I don't care if you're buying a sales tool, optimization software, or a CRM, whatever it might be. There are three reasons that you do it. As an executive you're increasing your profitability either by increasing revenue or reducing costs, you're improving customer service, or you’re reducing risk. In our case, we're trying to make safer routes too. If you're going to a dynamic routing option you make sure you're hitting those three things. If you don't hit all three. Well, then that would be the rare instance where you don't do it. I have yet to find a business that would not benefit.
At the end of the day, it's about absorbing that cost as they grow even if trucks have doubled in cost in the last decade. And there's so much room, so much space; the opportunity to get one more stop onto a route. And it ties into so many other opportunities where you change your driver pay program, look at productivity models, and reduce turnover. There are so many ancillary benefits. but it starts with absorbing that capacity and that cost. The companies that are doing this are growing. They are growing their profits because last-mile logistics is driving overall efficiency throughout the supply chain. I’ve lived this, that's what I did at Mohawk. And what we've seen too is over the years, right?
Bill Mathews: I have seen it, but it is so rare as to be notable. I’ve seen companies grow revenues by 25% and not add headcount or capacity.
Mat Witte: There are so many levers to pull for the executive and that's really what brought me back into this side of the business, talking about that at the executive level, and just opening up and that's right. That's why I think you're gonna be successful with this too. Because you've been around. You've seen it. You get it. It excites me to talk about it, and I can hear it in your voice as well.
Bill Mathews: I’m just convinced that companies can do so much more with their route operations. I believe there is always another 5-10% improvement out there to be gained with the right coaching and the right effort. We’ve seen what can be accomplished.
Mat Witte: I would sign up for 10-15% reductions all the time. Once you create the culture, all you have to do is pull the levers. It can take a while to get the culture right, but once that is accomplished, you have the setting to promote continuous improvement with the right tools.
Bill Mathews: Mat, it is always a pleasure to catch up with you
About ORTEC
For more than 40 years, ORTEC (ortec.com/en-us) has supported many of the world’s best-run organizations to make better data-driven decisions. Our decision support software and data science capabilities enable our customers to improve their business results and make a positive impact on the world. With our 1,000 employees across 13 countries, we help over 1,200 leading customers worldwide to make better choices in an ever-changing environment.
