In this transcribed episode of the Used Car Dealer Podcast, Zach talks with Dave Anderson, founder and CEO of EvenFlow AI, about how smarter shop loading and capacity optimization can help dealerships better match service demand with technician capacity while improving fixed ops profitability, employee productivity, customer retention, and used-vehicle reconditioning.
Dave brings a PhD in theoretical and applied mechanics, executive experience applying capacity optimization at United Airlines, and additional work in healthcare scheduling. He explains why automotive service departments too often operate like emergency rooms when they should be planned more like surgery centers, and why predicting likely demand is critical to using fixed technician capacity more effectively.
The conversation covers EvenFlow AI’s origin story, revenue optimization, technician capacity, burnout, used-car recon prioritization, recall readiness, customer choice, shop loading, advanced scheduling, integrations, and what dealers should expect during implementation. Dave also explains why AI can execute a solution at scale but cannot replace the dealership’s responsibility to understand and solve its own operating problems.
Whether you’re trying to unlock more production without adding headcount, improve recon speed, manage recalls more intelligently, or create a smoother service experience, Dave’s perspective offers a practical look at how dealers can better match fixed capacity with unpredictable demand.
Zach: Zach here, and today we're joined by Dave Anderson, founder and CEO of EvenFlow AI. Dave brings a PhD in theoretical and applied mechanics, executive experience applying capacity optimization at United Airlines, and additional work in healthcare scheduling, experience he is now using to help dealerships better match service demand with technician capacity. We'll explore what that means for service profitability, employee productivity, customer retention, and getting used vehicles frontline ready. So Dave, really appreciate you coming on the podcast today.
Dave: Yeah, thanks, Zach. Excited to be here.
Zach: So Dave, before entering automotive, you worked on complex capacity and optimization problems at United Airlines and in healthcare. What did those industries teach you about managing fixed capacity and unpredictable demand that dealerships have historically missed?
Dave: So, first of all, you know, I would say that I kind of stumbled upon this area, not that, you know, like, in a purposeful way. But what I, you know, what I learned is that, like, this is a pretty common problem that is unsolved for in most contexts. So, like, a hair salon has this problem, right? Like, some fixed capacity variable demand. Golf courses, movie theaters, right? Like, you know, airlines, obviously, with my background. So it's pretty ubiquitous out there. But it's not really well understood. And so one of the things that coming in with automotive and comparing it to healthcare, people often ask me, like, where's the parallels or whatever? I'd say the best parallel I'd say is that automotive wants to run, like, a surgery center.
But too often, it's, like, an emergency room. And what I mean by that, you know, emergency room is, like, it's, you know, it's almost like an oxymoron. You'd say, like, oh, it's an emergency. It should be fast. But the average or the median care time is, like, almost three hours, which doesn't seem so fast for an emergency, right? And the fundamental challenge with emergencies is that they're 100% walk-ins. Like, no one schedules an appointment to go to the emergency room. So you have all this, like, unpredictability of demand showing up. And then you have huge variability, right? If I go in for a broken arm or you go in for strep throat or someone has, like, a stroke, like, the resources needed and the skills is huge variability across that and variability. And that causes this problem, right? And then from a surgery center, it's almost the opposite in the way everyone has an appointment. So, like, you can't just walk up and say, hey, can you, you know, do surgery on my ACL? Like, you got to plan that out in advance. And then there's no, and it's a known procedure, right? Like, there's no notion of upsell, like, hey, Dr., while I'm here, you might as well take out my appendix, right? Like, you're there for one reason only. And so in that way, they can plan. You know, it's a pretty, having scheduled surgery centers, it's a pretty complex, you know, collection of resources that you have to bring to bear, but you know exactly what it is. And so you can pre-plan on it and they run super efficient. Then you take that in the context of, like, automotive, one of the challenges there for sure is that, like, all right, I have a vehicle diagnostic or a concern, right? Well, you don't know if that's going to be, like, 0.5 flag hours or, like, 10 flag hours, right?
You're going off of really impartial information here versus, like, a surgery center, like, you know it's an ACL, you know it's an appendix, you know, like, there's no, like, doubt about it. But, like, automotive, the challenge is you don't know it. So you have to try to, what we're trying to bring to bear is, like, do our best to predict what's the likely demand for that. So then we're planning, you know, with that prediction, so it's more like a surgery center, right? Like, if you just plan main shop as, like, a one-hour diag, and I scheduled 10 one-hour diags, you might have just taken in 30 hours of flag time. So we should plan that as, you know, 30 hours of flag time, right? If that's what the, you know, it's likely. So that's the real challenge with automotive is you're trying to predict what it's going to be so that then you can run it like a surgery center, which then is, like, very efficient. But the big thing about the surgery center that's different than what's traditionally, you know, in automotive is that, you know, the schedulers are appointment calendar fills, and they just jam things in the morning. So it would be like a surgery center saying, like, all patients show up at 8 a.m. And then we'll get to you eventually, right?
Like, it doesn't work that way, right? You got an 8 a.m. surgery. I got a noon, right? Someone else has got a three, and because they're staging it out, and that's where you get the most effective sort of, you know, care. And in automotive, that hasn't been part of the mindset, but that's where it's got to be in order for you to, like, the team to be set up for success and to fully service the vehicles and give the customers the best experience. So I know that was a long way in the way, but it's kind of like the, it's kind of like the challenges is too often like emergency room, but you got to run it like a surgery center, but you need some extra tools and know how to get it to be like a surgery center.
Zach: Totally. And that emergency center, surgery, you know, center is very compelling. But let me take a step back. So correct me if I'm wrong, EvenFlow's origin story began with you waiting three hours for an oil change. What did you see that day? What was the customer experience problem? Kind of give us a little more about the foundation of EvenFlow.
Dave: Yeah, and I'll, like someone asked me about this, and I just prefaced it of, I wasn't upset one bit because I was like, you know, I'm an entrepreneur. So if someone gives me free coffee and free Wi-Fi, I'm usually pretty happy. So I was just kind of chilling out, you know, in the waiting room. But, you know, at the end of it, you know, we were working on a health care company, and we thought we were pretty close to exiting. And, you know, we're kicking around the ideas of what next. And it just kind of struck me like, well, all I did was an oil change rotate. And I'm not a mechanic. I don't know what, like how long that takes, but it doesn't seem like it should be three hours. Like, and so it just seemed very inefficient. I saw a lot of people in the waiting room, you know, and they're kind of like me, you know, of didn't seem like they had that complicated of an issue. But we're all just kind of stacked up there. And it just felt like there probably was a better way here. But it was pretty naive, you know, approach. So then we studied it literally for two, three years, you know, kind of on the side of just trying to learn. And actually, I was very skeptical. I was like, oh, this is such a big space. I'm sure someone's already got this figured out. So it's like, so we'll just keep researching until we get this knockout condition to say like, oh, it's, you know, there's no opportunity. And we kept clearing toll gates until eventually I was like, oh, shoot, I guess we got a company, you know, maybe we should do it.
Zach: And how do you get to raise from Steve Greenfield, Automotive Ventures and FM Capital? How'd that come into play?
Dave: I was just really fortunate. I think I, you know, when early on, I didn't know anyone in the industry. And I just got hooked up with a couple of, you know, generous people who opened up their networks. Ron Fry, among one of them. And he introduced me to Steve. And then I think by then Chase as well. So both those firms. And, you know, I think when I first met him, I didn't even have a product. Like I might have had a clickable Figma or something like that. But I just, I think what we were talking about piqued their interest and we just stayed in touch. And then, you know, as we progressed, I think, you know, they got really excited about because there's so much about, you know, there's these AI scheduling agents and different, you know, tools out there. And they're really like a labor efficiency play, you know, engagement. And so there's no knock against that. There's real value to it. But no one was really touching upon like revenue optimization, right? Like in fixed ops, it's about technicians flagging hours and then parts follow on from that, right? And so it's, you know, it's really about maximizing the revenue. And so no one was tackling the problem that way. I think I just piqued their interest and I continue to kind of pique, you know, dealer's interest with that, you know, focusing on that pain point.
Zach: So shifting gears, NADA, they estimate an annual shortfall of roughly 37,000 trained dealership technicians. How much additional output can EvenFlow unlock before a store truly needs more headcount? And how do you prevent efficiency from becoming technician burnout or an unfair distribution of work and pay?
Dave: Yeah, so that's a great question. So first of all, what I would say, and I, a caveat is like everything depends on each individual service center. So, you know, these are, you know, blanket market conversations, but like in NADA's own data, the way I analyze it and then coupled with like about 150 dealers that I've looked at their data in detail, like the average dealership is maybe he's got at least 25 to 30% more production that they could get out of their technicians with the current staffing. And so there's a little bit of my United Airlines background where like we're always like on the verge of bankruptcy or out of money.
So the question was like, hey, I'm going to make you do more with less. So that's kind of like burned into me. So like with that kind of experience, I look at it and say, well, before you go out and hire a brand new technician, you know, why don't you see what you can get, you know, maximize what you already have, right? And I think most dealers that, you know, we work with, you know, there's always been the case where they've had like at least a 20% like opportunity for, you know, getting more with what they have right now. Now regarding the, you know, technician burnout, like it's actually, it's almost like the opposite. It's like in order to slow down, you can do more and speed up. So if you think about like where the burnout happens more often, it's due to the log jams, right? Like, so if you're a technician, say you're an express and someone comes back and says, hey, by the way, the promised time of this car is in 15 minutes and you need to get it done, right? Like if you got to do, because there's a log jam, then you're like, you're, and I've witnessed it, they're doing the fastest inspection possible, right? And they're banging it out and they're like, geez, you know, like I'm done. And then like the next one comes up and then, you know, maybe at three o'clock in the afternoon, they're dead, right? So that's a tough way of managing your day. Versus what we're trying to do is almost think of it as like create a conveyor belt of working at the pace that, you know, you can do your work. So it's almost like you complete a job, there's nothing backed up and then there's no downtime.
The next one comes in. So it's just an easier way of like going through your schedule, right? You think of your own day, you know, if I asked you to do eight things in an hour or 12 things in two hours, you'd probably say, I'll do 12 things in two hours. That seems a little bit better, right? And so that's the same thing. Now, I get it like, you know, some technicians might say like, well, I only want to flag a certain amount of hours, you know, or I'll take whatever I can get. Now, what we have the ability to do is really not only sort of the overall shop loading, but like kind of get it down to the individual level, almost like pre-dispatch or factor that in. So like if you're a go-getter hustler and you're just like, if I can flag 80 hours in a week, like give it to me. And I might be like, listen, I'm kind of just happy with 40, you know, like, and I do my 40 good and there's a place for me. We can factor that in and how we load the shop with that in mind. So it really just kind of, that gets into no two service centers are the same. And not a single service center is the same, every six months. They're like this living being that changes all the time.
Zach: Totally. And I wanted to move on to like used-car reconditioning and changing service specs. So Cox Automotive, they reported 47 days of used-vehicle supply in June of 2026, while average listing prices were 6% higher year over year. When inventory is expensive and everyday frontline matters, how should dealers connect service capacity decisions to reconditioning speed, inventory turn, and of course, front-end margin?
Dave: Yeah. So I think this is like, this is the holy grail, you know, and I'm getting into pieces of it, but we haven't quite fully solved the problem. But if you think about it, it's just like, you've got fixed resource, you know, and how do you maximize that time, you know, for, and used car could be, you know, if you've got your own used car, you know, technicians, all right, that's a one pool of resource. But if you're running it, we're like, you know, we're the main shop techs, but we also are supposed to do recon as well, right? Like now they're competing objectives here, right? And then you factor in CP versus warranty.
So ultimately what it should be is like, you're trying to maximize the value or revenue for each unit of time there. And so factoring in like, when should recon be pulled up? Or when should this, you know, warranty job or CP be pulled up? And how do we balance that? I think it's a pretty dynamic problem, but that's how you got to think about it. But it's also like, the way I look at it, it's not all recon is created equal, right? And it's also, that's another, this is one of the weirdest things about the, and it's fascinating to me as an outsider of like, with the optimization of recon, because like in most places, oh, I can, it's hard to think of other contexts, but like what you have on your shelf to sell is a deliberate choice. Like, you know, you're stocking things, right? You know, used cars is random, right? I don't know who's going to trade in what car. You can look at historical patterns, but like, you know, if you're a Honda dealer, you might also get six civics in a row, right? And no odysseys. And you're like, how did that happen? Right.
But so it's this random, you know, input on what you're stocking. But when I look at that, a part of the problem is like, we can't create it all the same, right? Like, so if you get six civics in a row, that six civic, maybe you don't need to recon that one so fast, right? Because you already got five on the lot. But the next odyssey that comes in, man, you better bump that up in a line because that'll probably move fast, you know? Or if you're a Honda and you got a Mercedes, like it's a beautiful car, but you're like, I'm probably just going to have to wholesale this. And it's right like, so it's trying to be smart about how you prioritize that inventory of how fast it will move and then balance it off of the other competing, you know, time of the technicians. Of this warranty job or CP job. And that's the holy grail of how I would look at it. Right now we kind of do it crudely, kind of bring it all together. It's like some math is better than no math, right? But it's not like sophisticated, optimized. Like what we do on CP warranty, like we're pretty robust. Like how we pull in recon right now, it's just, I call it some math, you know?
Zach: So one of our portfolio companies plays in the recall space, The Recall Company, they were formerly known as AutoAp. How does EvenFlow kind of work in the recall landscape? Or how does it touch, you know, the recalls as far as service appointments?
Dave: Yeah, yeah, yeah. No, and The Recall Company, they're great partners in what we're trying to do here. So I look at it as sort of four facets with recall. And just from an operational standpoint, I think of it as like customer readiness. And what that means is, does the customer know that they have a recall when they're scheduling appointments? So they're not showing up in the drive and caught off guard of like, hey, I just had an oil change and now you're telling me I got these, you know, three other things I need to do. So you got to make all booking channels, like, you know, make sure you check and disclose the recall. And then another part of readiness is communication. So it's, some of them, there's like policies or procedures, right? Like, hey, Zach, this is an inspection first. And if this fails, then this is what's going to happen. So you should disclose that ahead of time. No surprises, right? Or this recall is like a software patch that's, you know, no big deal.
This one, you're going to have to like, you should, we recommend you drop off your car because it's going to take six hours to complete, right? Like, again, no surprises in the drive here. So that step one is customer readiness. Step two is operational readiness. So not only is it technician capacity, but it, do you have the parts on hand, right? Do you need, you know, lead time to order the parts or are you on a daily allocation and you can only do five a day and this is the sixth request. So we better move it to the next day, right? So, and then other communications, the parts managers make sure that they check that things are on stock, in stock, right?So that operational readiness is critical so that like the customer could be ready, but then you're ready to pull it through, you know, in the drive there. And then third is business optimization, I call it, which is kind of like along the lines of what we were talking with recon as well, because, you know, not all recalls pay, you know, equal, right? And so some are less attractive and if that's what you're scheduling it for, then you kind of push it to, you know, what I always recommend, push it to like off peak periods where you have softer demand and like free up, you know, the kind of creating that better flow in the peak periods, think of it like in the morning and then you kind of turn some of those nuisance ROs into, you know, opportunities for CP, you know, and cross sell there. And this one I learned from healthcare because, you know, we used to talk to doctors, I could say this now because I'm outside of healthcare, but they'd be like, I don't want these Medicare patients, like, ah, these Medicare patients, they don't pay anything. Give me Blue Cross Blue Shield.
So, you know, it's like, ah, yeah, I get that. But, you know, doctor, you can never fill up that 2:30 slot. Like, why don't we just push them there? Like, and they're probably retired. So they probably have flexibility and time. Yeah, OK, we'll do that. So it, you know, it's like that business optimization of like, you know, matching that fixed capacity to the value of that time, right? Of what you can generate from that. And then the fourth piece is really like more of, like, you know, the recall alerts of being proactive, of pulling, you know, notifying customers that they have the recall and pulling them in to schedule it with all the customer readiness, you know, you know, operational readiness and business optimization. And there, though, it's like we're still learning. But I think a big part of it is, first of all, I would say, you know, it's not quote marketing, but from a marketer, like it's a perfect thing. It's like, hey, Zach, this is this is for your safety and it's free. Like there's no better like messages you could tell someone, you know, with it other than like win a million dollars. But, you know, it's like it's it's for your safety and it's free. But then there's things in the message that you can make it like feel safe and secure to the customer, right? Like here's the NHTSA campaign. Here's a link to NHTSA. So we're for real, right? Like you can go read it, you know, make it easy for them to research the importance of it, right? Like so all that is like key of pulling customers in to do it. But it's got to be within that, like, you know, operational customer readiness mindset.
Zach: So totally. And our friends at Automotive Ventures, they recently highlighted advanced scheduling capabilities and they mentioned open recall capacity management, employee assignment algorithms, dynamic maintenance pricing, along with possible applications beyond service scheduling. Which capability is most under appreciated today, in your opinion? And where must dealers draw the line? So optimization never undermines transparency or customer trust.
Dave: Yeah. I think. So I'll answer the last question first. This comes up a lot. So the optimization, it still has got to be done with choice, right? And one of the things it's almost counterintuitive to people, but we usually, when we come in, like beforehand, a dealership might have like 15% like same day, next day appointments. That means like I'm sitting here on like a Tuesday morning. I'm trying to get in Tuesday or Wednesday, right? We generally track to like 50%. So we significantly improve it. And it's kind of counterintuitive to say, Dave, you're putting in all these controls. I would think it would push it out. But there's no bias in how we're setting this. It's just truly available. And so it puts choice back into the customer. So all the flexibility is we're trying to achieve something operationally for the dealership when it puts choice in there. So like, do you want to come in Tuesday afternoon or Wednesday morning, right? Do you want to loan a car or do you want to wait, right? All these things. Do you want to come in at in the afternoon and save $20 or would you rather come in in the morning and pay sort of the full price? But you put the choice in the customer's hand. And so in all industries I've worked in it, as long as you let the customer choose, you know, they generally won't be. They're not going to be upset, right? They made the choice.
He gave them A or B. It was their choice. You know, there's no constraints there. So in that way, it's critical to like deploy the optimization and like where the consumers have lots of choice. For from there, I'd say, you know, the thing that's most underappreciated still today of it's just the course, you know, shoploading capability of understanding it. Because so many, you know, people have grown up in the industry of like, I'm just filling calendar slots and not even, you know, thinking from the production capacity perspective, right? Like if I got 100 hours of production capacity, how do I maximize that? Like fill things up. And that means maybe I don't take in 23 cars at 8 a.m., right? Like I stayed across the day there. But the thing about that too is that always increases customer experience. Because now what you have is like you can, you know, the advisors can actually stand behind their like promise times, right, and set the expectations. So like, you know, Zach, we can get you in and out in an hour. Like I'm commenting, we're like, hey, this is going to take two hours, right? You know, before it's like, well, we think we can get you in maybe an hour, hour and a half, but we'll let you know when it turns out to be three hours. Now you're just waiting for someone to stew and be upset, right? So it's that, you know, no surprises, right? And that's what we help the team be able to do.
Zach: And EvenFlow's site, they list integrations across a range of DMS and dealer-tech platforms for an independent used car dealer or smaller group operating with a mixed technology stack. What data is truly required to get started? And what should management expect during the first 30, 60, 90 days with EvenFlow?
Dave: Yeah, so from an integration standpoint, you know, so I'm still trying to get my feet wet on, understanding all the independent different software systems out there. Like really like what we, the most important thing is making sure that we're pushing in so that they can see it within their point of sale, right? So that they know that, you know, Dave is coming in at 8 a.m. with this vehicle, right? So it's integrated in their system, you know, for that, that they can pull it through once that customer shows up. That's the most important aspect of it. You know, from how the change goes, like it depends on the nature of the dynamic of the shop, right? And so, you know, faster turning, you know, brands like Honda, Toyotas, right? Like where generally like, you know, what comes in a day comes out at the end of the day, right? Versus something like CDJR has got, maybe it could have like a three-week backlog, four-week backlog of like repairs, right? So it depends on that context. But like, if you're in that, you know, the Honda, Toyota, really by like week two, you should be seeing something, feeling something different, you know, already.
If it's a CDJR type store, you got a bigger backlog, depending on your nature, like it could take like three weeks, four weeks before, you know, you know, you start feeling something materially different. But that's why I always ask, like, you know, besides normal shopper, like what's at least one key business objective you want to achieve like immediately? And then we really try to like, go all in on making sure that they see some benefit earlier on of solving that particular pain point. And then we can build from there. The one thing I would always say is a watch out though, is you can't just tinker with it. Like it can't, it's not like the volume control on your TV and just always change it. I got up to 17. Let me turn it down to 14. Let me turn it to, you know, 13.
Let me turn it to 21. Like you gotta, you have to let some things kind of like set for a little bit or else like you just get like chaos of all the settings, you know, moving around there. And I think it's a, I get it. It's a little leap of faith. And I always tell everyone like at some point I'm going to mess you up. If I knew what that was, I wouldn't mess you up. Like, but I'll probably mess you up and then call me and then we'll recalibrate it and we'll get it right. That's part of the onboarding process.
Zach: So well put. And the last two questions, this is a rapid fire round or sorry, three questions. So I'm going to ask you three rapid fire questions before we close out the podcast today. And Dave really enjoyed everything so far. So number one, what is the one fixed ops KPI every dealer principal or GM should review daily?
Dave: I would say flag hours versus production capacity, because everything else, like, service growth follows from there. But, like, that's your leading indicator. Are you actually, like, you know, producing from your, you know, your revenue producing capacity? So you could flag 100. What did you do? You only flagged 82. That's a problem, right? I'd ask a question then.
Zach: Number two, what is one dealership habit or management problem that no amount of AI can fix?
Dave: Ah, I would say almost all of them. So here's the thing that, and this is a bigger question about AI, but like, like, AI is amazing, but it's not going to solve a problem that it doesn't know, right? Like, it can execute, you can train it on a workflow, you can train it on a decision, like how to do it at scale. But like, so for instance, like even the recon example, you couldn't just spin up AI and somehow magically figure out how to optimize recon, but you could come up with like a framework or the math behind it. And you can then get AI to execute it at hyperscale, you know, 24/7. So to me, that's the biggest thing that I would tell everyone with AI is like, you can't outsource the solution to your problems through AI, like you still need to own your, your solution. And then AI is a tool to execute against it. So that's where I just say like all the business intelligence expertise and running your operation, you can't let go of that. And then you know, that gets fed into AI, not the other way around, if that makes sense.
Zach: Definitely. And lastly, what's your favorite AI use case in your personal life? So outside of business, your own company's tech, anything like interesting for the audience?
Dave: So I still use this, this still kind of bleeds into, you know, my personal life, but I mean, my business life, but because it's so hard to separate out my, because as you know, as an entrepreneur mode, yeah, like, where's the line, right? Like, so there's like strategic questions that I have. And so I might start a chat with say, like Claude and like having a conversation, but then I start asking like my wife, or if I ask an investor or ask an employee, I'm like, hey, pick up on this thread and then go ask, you know, chat GPT or your own like conversation with Claude, and just start going at it. And then I like then getting all the threads back, because I don't know what prompts or questions other people will ask, or they see a problem a little bit different, or they see, you know, the LLM, some different pieces of information. So then if, but if I can get five conversations to come back, then it's really interesting to see with me what is kind of a truism, like what always comes back no matter what, right? And then where do these things deviate? And if they deviate, what made it deviate, but that like is incredibly informative. I just started doing this. And it was like, it was like, wow, like this is and not doesn't necessarily even mean that you get like, I saw even a new insight, but it like really crystallized like, hey, this thing is true. Like I can put, you know, I that's rock solid, true. And this one, maybe that's squishy, I got to learn more about, you know,
Zach: very well put Dave, and loved having you on the podcast. For anyone wondering how to contact you get ahold of EvenFlow, how should they go about that?
Dave: They can go to the website. It's evenflow.ai. Or you can always email me. It's my name Dave A. as in Anderson, but davea@evenflow.ai. And yeah, drop us a line. Love to love to connect.
Zach: Awesome. Well, thanks for being on the podcast.
Dave: Yeah, thanks for having me.
.png)
.png)
.png)
.png)
