Latest / Exit Algorithms / Cut Excess Inventory, Unlock Cash Flow & Forecast Smarter with Alex Maibaum (#37)
Transcript
- speaker-0: Welcome to exit algorithms, the podcast where we decode what it really takes to unlock growth, streamline operations and prepare your business for a high value exit. I'm your host, Pete Vera. And today we got Alex Maybaum in the building. He's a supply chain. sir. He's a supply chain veteran, tons of experience in the founder and CPO CEO ⁓ of predictive chain solutions. a cloud-based software consulting company that helps mid-size businesses forecast smarter, cut excess inventory, and unlock some serious cashflow. Alex, super stoked to have you. Welcome to the show. speaker-1: Thanks a lot for having me on today. It's a pleasure to meet you. speaker-0: Yeah, absolutely. I'm excited to get into this. Do you mind, we can start off to share a bit about your background, for you to go as far back as you want. then, yeah, just sort of your business journey. speaker-1: Absolutely. Yeah. So I, I've kind of worked my way from the ground up. it was really, you I've been really fortunate. I've been introduced to a lot of businesses over the course of my life. and I've had a lot of opportunities in different parts of businesses. And I've always found that to be really helpful, whether I'm doing sales marketing, working on a warehouse floor, every single piece of the business you touch, you kind of learn a little something new. What's going on in that? What's the nuance of that particular department? All those things kind of fed. the interest throughout my career to understand how businesses are functioning, where the inefficiencies are located, and how they can continue to function better. Probably the biggest unlock, though, I realized over the course of the time of working in these different businesses is how frequent it is that these departments really don't communicate with each other. Sometimes they do communicate, but they don't communicate effectively. And I think a lot of that is really driven by the fact that the ways that we communicate between businesses and between departments can sometimes be coded. We speak in the language of that department. And that can sometimes create barriers when you're trying to communicate and make decisions across the business. This was a really big unlock I had when I worked for the company SpyOptic, a little Southern California based sunglass company, not little, they're a pretty big Southern California based sunglass company. And what I started to notice was, that as I was doing data analytics for them, the information they were trying to get was there. It was there all the time. What we need to do is decode it so that when the sales team was communicating with the marketing team, communicating with the operations team, everybody was speaking that same language. We used the application Power BI there to actually build out a reporting environment for them. And it was really fascinating watching. how once the data was together, everybody agreed this is our source of truth. This is where the information is coming from. How quickly this, I'm gonna use the word synergy here. I know it's the buzzword, but that is really what happened. You started seeing synergy across these departments. People were all talking about how can we optimize what the business is doing. And we started to see some, lot of unlocks. I took that with me to a couple of other businesses after that. One of the areas and kind of the genesis of what I realized was the one of the biggest gaps in businesses was the disconnect between sales and marketing operations and finance and executive teams. The analogy I love using is it's like a ping pong table. You've got sales and marketing trying to do what they need to do on their side of the ping pong table. And you've got a finance and executive team on the other side trying to keep the business moving. And what's happening is the ball is just going back and forth all day long between those two teams. Kind of in the middle there, you've got operations. We're kind of sticking our hand up every once in a while, trying to grab that ball and try and make sense of what the business is doing. And what you have here is, again, you don't have a synergy, you don't have synergy across these departments. And it makes sense when you think about it. Finance is speaking in very specific finance terms, works really closely with executive teams to make sure that the business is going in the direction that they're trying to go. Then you got sales and marketing teams are extremely agile. They're moving quickly. They're reacting to what's happening in the market. two things don't always communicate well with each other. And then operations is trying to figure out how do you feed the engine? know, I'm feeding a coal today, wood tomorrow, ⁓ electricity, you name it. We're trying to keep everything moving, but the reality is that the business shift from finance and executive teams, we don't have enough cash. Sales and marketing team, I need more of this. the timeline on those things doesn't line up. ⁓ I realized is again, this became a data visualization challenge is really what it came down to. How can I give businesses the information they need to make better decisions in a structured manner, remove those barriers of communication? The way I did that was laying out the data and it's what we do at Predictive SCS today. I laid out the data so that the sales and marketing team can see what their trajectories are. operations can see how that impacts the business and then finance can provide their input for this is how much we have to spend. What that does is it provides better operational clarity to businesses and that's really the key ⁓ unlock to a lot of these businesses. You're no longer reacting to what's happening. Sales and marketing has a new opportunity. They're going for it. Now you're reacting to what's happening. Leveraging a system like this allows you to be much more proactive. if things change. sales and marketing accelerates. You're able to see that moving within the business. You're able to see that in the form of your inventory and your future inventory positions. It allows you to make much more proactive strategic decisions. Now for our conversation, what we're talking about is that translates immediately into operational cashflow efficiency. If you're not buying too much inventory, your business is turning inventory faster. That's your business term of the day there. Inventory turns. means is that you're converting inventory into cash faster. For your business, that means you have more cash on hand to do more things with it. So that's how we got to where we are today. And it's been really exciting watching a lot of businesses we work with grow as a result. speaker-0: Yeah, that's awesome. I love that analogy you used. I've heard a few. There's the ball, the game one, and also kind of an engine. It feels you're, so you're kind of making ⁓ it for all the parts of the business to communicate. You're like oil in the machine. speaker-1: That's correct. Yeah, I think again, the functional disconnect here is if businesses don't have a way to, yeah, if there's no lubricant to allow these businesses to communicate between each of the different departments, you have a functional breakdown. I was just talking to somebody recently and ⁓ one of things somebody mentioned to me is it's like car is another great analogy for a business. Most people understand generally how a car works. There's an engine in there, it consumes fuel and that makes the car go. When you open up the hood of the car though and you start to look at the engine, there's a ton of components. That is your business. You have finance running, your accounting department, you have your sales teams working in the field, you have your operations, you have your warehouse team managing the product every single day. All those components work together. If they're not lubricated, things start breaking really quickly if you're not doing the maintenance on them. And that's often where you start to see these disconnects happen. speaker-0: Yeah, no, that's awesome. I like that car analogy. I've heard the plane analogy. Yeah, with each one wing is sales, one wing's marketing, the fuselage in the middle is the overhead. speaker-1: Yeah, exactly right. Yeah, it's really fascinating how quickly you can see the change within a business you actually start to level all those things out because your team moving from the constant feeling of firefighting, ⁓ my gosh, we're out of stock on something or we're really low on stock, we didn't place an order for it ⁓ or didn't realize ⁓ that this was going to accelerate like this. Last month when we looked at this, it seemed like we had plenty of stock. this month all of a sudden it seems like we're really low on stock or, you know, finance is saying, you know, Hey, the business wants $200,000 worth of buy dollars to go buy new stock of product and finance is going, yeah, we can only afford a hundred thousand. How do you bridge those gaps between those challenges within a business? Because you know, as a business owner, there's opportunity. So if you had 200,000, you could sell all that inventory. but now you're stuck in this position of saying I have to constrain the business. What we really want to do is we want to unlock these businesses to be able to say, no, we can get to that 200,000 of buying, but we're going to do it strategically in a way that's actually going to make that inventory more impactful for the business. speaker-0: Hmm. Yeah. ⁓ I love that. Yeah. And you mentioned reactive versus proactive. Is that kind of the probably the make the biggest, would you say value add is being able to ⁓ ⁓ get ahead of things? ⁓ speaker-1: I don't think it's necessarily so much about predicting it so much as it's about trying to understand and figure out is proactive, but breaking that down a little bit further, it's about seeing what your business is actually doing today and understanding what's working, what's not working. In is especially within operations, there's the 80-20 rule. 80 % of your sales comes from 20 % of your product. And a lot of businesses I see it's even smaller than It's like 90-10. You got 90 % of your revenue coming from 10 % of your products. So what we see a lot is skew bloat, where businesses bring in more more more products trying to chase more revenue, but they don't understand how much revenue is actually being generated by that product they're bringing in. But then they experience this compounding challenge of, now we brought this product and we need to keep it in stock because we have a customer that's buying it, and so we want to make sure it's available for those customers. And so one of the things that we work on routinely with these customers is better understanding what's actually driving the revenue to the business. So this is not to say like you can't go buy new things. The idea is to limit the scope of what you're buying so you're not exposing yourself to excess risks. And when we talk about being proactive, we're talking about working with your factories. Can you lower your minimum order quantities, MOQs? and you get better payment terms and things like that that can ⁓ the strain in your cash flow. Keep those products available because we always want to be injecting new products to our customers. We just don't want to get to this point where we have this massive skew portfolio of products that's driving very little revenue and putting enormous cash strain on the business. speaker-0: Interesting. Yeah, that's a cool application of the Pareto principle, the 80-20. that probably the biggest issue ⁓ see, you find with the businesses you work with? speaker-1: Yeah, most of the businesses that we're working with, have two compounding issues. You're to have that 80-20 rule where you have a mismatch in the number of SKUs. ⁓ then what we're finding is a significant overstock in products that just aren't moving product in consistent manner. If you want to get complex, we get into the difference between statistical analysis, demand planning. That's really the of ⁓ how businesses are understanding what they need, not just today, but tomorrow and in the future. That becomes really important for businesses to better predict what they need. And that will drive a lot of that proactive nature in the businesses. It starts with trying to figure out, answer your question again, we've got too much inventory. Yes, that's a common thing that we see in these businesses, too much, too many products. too much inventory across spread across all those products. So what do you start to do? We start to find ways that we can trim the products that really don't make the most sense to the business. Let's actually put a dollar to that. If you've got a million dollar business and you've got the majority of your business is being run on that 80-20 rule, that means you've got 800K of revenue coming from 20 % of your SKUs. The breakdown in eating math, you've got 100 SKUs. That means 20 of those SKUs are driving 800K of your revenue. And you've got this other 80 that are just kind of languishing there. But if you actually go and you look at, start looking at those SKUs, maybe you're holding 100 widgets of one of those that are in that lower category there, but you're only selling one a month. So you have very limited revenue that's coming from a SKU that's holding a ton of stock. But if that's eating up your balance sheet because you are now stocking that product, it's taking a long time for you to get through it. The analysis here is understanding, well, I can't buy less because I have a minimum order quantity with my factory. best way to start looking at this is understand what is your carrying cost to that product. So if you have to buy the minimum, how long is it gonna take you to move it? It's gonna take you over a year. That's a lot of time to be holding the stock of that product. Then you start understanding what's the cost of holding that. Think about your opportunity cost. If I bought 10 of those instead of 100, what could I buy 80 or 90 of another product that maybe you're turning over four, five, six times a year? That becomes your cash unlock. speaker-0: Yeah, wow, powerful. I love that. Yeah, ⁓ want to dig in a little bit further on the data that you use, your software, maybe in ⁓ layman's if you describe your process a little bit, ⁓ what numbers use and sort of how you draw conclusions from the data. speaker-1: Yeah, and I think one of the most powerful things that I coach businesses on is there's a level of complexity that you can go through. What we're doing is can be somewhat advanced and there's a tremendous value application to businesses that do adopt this. But I think it's also important for business to realize there's a step before that that a lot of businesses can take and there's a lot of opportunities there. There's tons of tools out there today that can help you build rudimentary versions of a supply and demand plan. And when you start to leverage those, if you understand what you're looking at, they can be a very big unlock for your business. Again, driving towards that proactive nature of planning instead of reactive of, my gosh, we're out of this product today, we need to order tomorrow. So that is base to build something like this out. And think this is important because the goal here is to help these businesses grow, right? So I can pitch my software all day long, but it doesn't help a lot of these businesses need to grow. So the real idea here is how do we help these businesses to actually leverage tools that are out there for whatever stage of the business they're at. If they can continue to grow, there's always opportunity for these businesses to continue at that point. So the first thing, there's four key components to building out a successful supply plan. You need to understand your inventory positions. So establish a source of truth. If you're a business that doesn't have a good source of truth, start there. Really dive into that. Is that Shopify? Do you have a warehouse that you're paying at 3PL? Can you get a report from that 3PL? Can you get it scheduled so it's automated? Leverage a tool if you're not familiar with Excel to populate into a static space that has your inventory information. Not uncommon. Almost every single day I run into businesses that have zero system at all. And so the first thing we do is literally build a structured environment for them where we can say this is your source of truth. This is what your current inventory position is. Once you have your inventory, the next thing you want to do, consolidate all your purchasing. lot of businesses are maybe using QuickBooks where they're entering their bills and everything, but they may not have a consolidated purchasing system. ERPs exist, but ERPs are very expensive. So if you don't have a good consolidated view of everything you're buying, Just build out a simple Excel spreadsheet. What is it you bought? How much did you buy? What price did you pay for it? When is it expected to come in? Those are the key pieces of information. If you know that, you can continue to kind of evolve the way you're looking at what's going on. You've got your inventory and your purchasing. Next is understand what your historical sales are. So if you're in Shopify, you can run a really simple report. what your SKUs are, ⁓ individual product. ⁓ you know, use UPC or skew then run your last 12 months. If you've got 24 months of data, even better. If you trust the ⁓ AI models that are out there, you can literally take that data loaded into an AI model and say, hey, can you populate out? Like, is there any trend analysis or anything on any of these items? That's what we would do as our first ⁓ We go a more complex than that. Obviously we have some proprietary algorithms that we use, but you can go that first simple step of understanding like, hey, here's what I did in the last 12, 24 months. What do the next 12 months look like? And that can actually serve two purposes. You can take that information, you can apply it into a financial model now. That can be really powerful because there's frequently a disconnect between, again, finance and executive teams saying this is what we want the business to do what the actual business is doing. And sometimes that can be a huge unlock. We thought the business was growing at 20 or 30 % or we want the business to grow at 20 or 30%. But right now we're only seeing 5%. Where is that disconnect coming from? Is it because we think we're going to be releasing some products in the second half of the year and that's going to drive growth? But that's great. This is how we start talking about being proactive in nature. So now you've got your inventory, your purchasing and your sales, your historical sales. The last piece to this that we will really unlock is understanding if you have big commitments that are out in the future. This is one that stumbles up a lot of businesses. If you're starting to expand your business and you start to work with large retail chains, for example, Walmart, Target, Kroger, or it could be something like a Dick's Sporting Goods, a Mac Tools or something like that. They often place very, very large orders. And if you don't log those, you're going to miss the opportunity if you don't start planning for them. So see this routinely where somebody will come in and they'll place a really large order and you know, the business gets really excited, right? It's like, yeah, we logged in, you know, that million dollar sale. And then ops is going, ⁓ my gosh, we just blew out 99 % of our inventory like, and it's going to take us four months to replenish. You just yourself in the foot, right? So getting that last piece of information and always having that in there. becomes really important because now the ops team is correctly planning in the future, recognizing that there's a big opportunity. But this also goes into exactly why we want to build this all out. Because now if we connect all these things together, your inventory, your future purchasing, those two in particular, now when you do have that opportunity, you have the data in front of you to be able to say, can we take advantage of this opportunity? just locked in a big deals with exporting goods, we're going to sell into 150 of their stores. It's going to take a thousand units of inventory. Well, I've only got 1400 units of inventory on hand and it takes me six months to replenish product. It's coming from Southeast Asia. That could be a problem right away. so understanding that and having that visibility helps the business to make a much more strategic decision. So it really boils down to getting that data first. figuring out what it is, and that's a core area where we will help a lot of businesses at first. Let's figure out what your data is. Let's understand that. Let's build that source of truth. Then we can start looking at proactively, what does your business need to do? Let's work on the execution plan after that. speaker-0: Got it. I appreciate that. Yeah, that answer really good layout of the process and really good takeaways as well. Yeah, on the topic of technology, you AI a little bit. It's hard not to mention AI these days, but I'm interested in your perspective, being in supply chain data, ⁓ using technology every speaker-1: Awesome. speaker-0: as a core part of your business, how is AI impacting SCS or how are you guys leveraging it? speaker-1: Yeah, for us, one of the biggest unlocks is a levered unit on the demand planning forecasting side. Traditionally, this would be in the role of a data scientist. It would kind of restricted to much larger businesses because data scientists are expensive. ⁓ It's a specialized course of work. You've got very, very mathematical equations, models, and things like that you can apply. So it was kind of always segmented off to the side is like if you can afford it, it's awesome. Now that that entire space is opened up. I think the key thing to understand though is just because you can generate something doesn't mean it's correct. And so there's two big things that we really focus on when we really ⁓ we leverage AI. We use it all the time every single day. But when we plug in information, there's two things to consider one. understand the security level. The free version of all the large language models ⁓ open source. So if you plug your product costing information into there, the language model is going to retain that information. So you got to be aware of that. ⁓ If you're paying for some of them, at least today, they're telling you they're not training it ⁓ on your But so that's that's a key component to be aware of is, you know, just making sure that you're secure with the information that you're giving the model. The second is validating the information that comes from the output. It's really easy to build what would be called a black box ⁓ a model. That hey, here's my sales information, give me a forecast. If you don't understand how the model built the forecast, you effectively have a black box. You can't ⁓ go through the math and understand what it generated. Even if it built out, Claude is a great ⁓ use case for this, can build out a very can do things like layering in Python models now and things like that. Really, really advanced, really cool. This is stuff that I would have taken. Hours, weeks, months building out in my you know early in my career. Now we've done in less than a minute. That's really awesome, but you need to make sure that you have somebody on staff that can translate what it's actually doing because then you want to make sure you can validate it. Biggest. roadblock I see businesses run into, we created something we can't update it. So I spent all this time building this really cool AI forecast model or the supply chain model, but we functionally don't know how to rebuild it or keep it going. And so you got a one-time snapshot. Somebody spent a ton of time probably going back and forth, iterating on it, but you don't have a great way to actually update the information. So now you're in a position where you're having to go back through and iterate again. And you kind of you lose a little bit of that core understanding of what's actually happening with the business. speaker-0: Yeah, wow, good warnings too. Yeah, it makes sense. Black box. I haven't heard that term used in this context. Smart though, it makes perfect sense. speaker-1: I think that's a really critical one because again, used to do this all the time, right? Like they build something, but then they basically lock what they built. Like that's their IP. And in the consulting room, we call that the box. And what want to avoid now is building those things, but the AI models can be notorious for accidentally doing it. They're not doing it nefariously. They're just building things for you. without giving you the background information on how it was built or what was built to, or what it's going to do, how all the different pieces. There's a great tip. If you do build something out, go back to the LLM at the end of your build and tell it to give you a wiki. me a wiki now for what you just built. And will usually, it should go through and it should create either a doc, if you're using Claude, it will create a document for you. If he's using Gemini, it'll give you a write-up. It'll tell you exactly what formulas were used, what those formulas do. ⁓ how they function, that can be the key. Because now you don't just understand that something was built, you understand that exactly how it works, how you've gotten a 300 page binder now. It tells you the manual for what everything is, but at least it's not lost. And you can empower yourself to actually go in. What becomes really exciting then is if you start to read into what these things are doing and how they're generating this information. You can actually go back into your chat now and you can re prompt. Hey, this piece doesn't seem like it's working. Can we reformulate this piece right here because something seems to be outputting that I'm not expecting I'm doing demand planning. As an example, there's nuance to this particular business. It seems like they have a sales spike ⁓ in other year in this month ⁓ because doing a promotion. Cool. It can go back through and it can iterate on our demand plan model to ⁓ build in the nuance specifically for that, but I'm able to pull a specific part of the formula out and say, hey, I think we need to modify this piece because it's not interacting correctly with this business. But I'm only able to do that because I'm able to get the wikis out when I'm building these things. speaker-0: Wow, yeah, excellent advice. Yeah. think especially relevant for people that are building with AI. Do ⁓ use and have any other use cases maybe even in your personal life or other ⁓ applications? speaker-1: I mean, I'm using it all the time now from, I love to cook, so I'll use it when I validate recipes. I'll like, yeah, I don't have this particular ingredient. I'm thinking of using this ingredient. Can I use this instead? Yeah, but you're gonna create a sweeter flavor by doing that. So balance it out using this thing instead. Yeah, there's tons of applications outside of business. I've built my workout plans using AI now, which has been a lot of fun. But yeah, I think the applications here, I've heard this before from somebody else, the application of AI can be almost overwhelming. And the best way is to just kind of get started and do something simple. What do you enjoy doing? And then just start tinkering in that space. Because it's that space where you're functionally knowledgeable. And so you can, ⁓ the you have becomes more positive. as a result of that. love cooking, so when I'm cooking and I'm interacting with the AI, I can see, oh, that doesn't seem correct, or I can understand how that would work. So I'm able to validate what I'm talking about or what I'm interacting with. Whereas if you go and you do something like finance, and you're not a finance person, it spits out this massive detail. It's like, oh my gosh, this is an MBA crash course right here, and I'm not ready for an MBA crash course in finance. speaker-0: Yeah, definitely. That's 100 % true. Yeah, it's as useful as you make it. feel, you your inputs will dictate how useful it is. a lot of people, some people are getting fed up with the hallucinations or just like how sometimes it just spits out crap that's not true or whatever. But I feel like it's important to realize humans do that ⁓ know? And the end of the day, can, if you ask it to generate questions for you, like if you make it more of a dialogue, it becomes a lot more useful in it. It learns your habits and way you talk and. speaker-1: That's why I say like get that wiki so that you understand not just that it created something for you, but how it created it for you. Because then if it does hallucinate, you're going to see that in the wiki. going to wait a second. That doesn't make any sense at all. Again, this is more applicable to when you're running Excel type stuff or where you have complex formulas ⁓ or mathematical equations. It's definitely more applicable in that space. But there's equal use cases, I'm sure, outside where when I'm cooking, for example, if it makes a recommendation for me, I'll sometimes ask it to link me to a chef that is using that particular method. So now I can cross-check and understand, like, yeah, does that actually make sense? When I'm using it in demand planning, there's times where it will, I'm by no means a professional statistician or anything like that. I'm familiar enough with the numbers and how things are working. I understand now how to build out demand plans. But again, this was an MBA, PhD level course of work in statistical analysis to build these things out. And so there's times where it'll give me really, really complex methods of analysis. And I'll tell it like, hey, can you link me to an article that I can go read to help me understand what this is that you're building? Like what is this method of analysis that you're building out for me? ⁓ So far as I know, I I think it's working, but again, this is way beyond my pay grade, if you will. So I need to understand exactly what we're talking about. And then I have to decide how deep I want to go down the rabbit hole myself. But at least I know it's not just hallucinating. It's giving me something where I can go and I can actually read, you know, some PhD paper or something like that. Okay. Now I understand what Crosston's method is. Great. Never heard that in my college degree, or in my college course, but at least I have a way to go validate that it's not just pulling something out of the rabbit, out of the hat. speaker-0: Definitely, that's smart. Make sure what it's saying is true. I'm trusting on face value. But yeah, it's been a great conversation. know we're wrapping up here on time. Cool. But I just want to give you the floor, Alex, any final words? Where can we find you? SCS and if you any last word advice for listeners. speaker-1: Yeah, exactly. Yeah, absolutely. Thanks. I really appreciate it. Predictive supply chain solutions. ⁓ can find us at predictivescs.com. Our goal is really to help businesses grow. ⁓ That's passion. come from, we are operators, functionally we're operators and we've come from a lot of high growth, ⁓ hyper accelerating businesses we've seen what type of solution can do for businesses. It's functionally why we wrote the platform that we wrote. it was missing for the businesses we were working in. And so we wanted to make something that we could actually leverage to help other businesses continue to grow. And so we're really excited to be continuing to work with businesses. do continue expand the people that we talk to. Our biggest hope is that we can be a resource to as many businesses as possible. reach out to Let us know if have questions. We're happy to chat. One of the things we love is the fact that we are still functioning operators. As much as we're building a software solution, the goal is actually to continue to be operators. I think one of the biggest challenges that exists in the market today is there's a ton of solutions out there and a lot of them are written from a software and engineer's perspective. And that disconnects sometimes from what the user is actually trying to do. So you have this paradox of, there's really cool software is out here, but it almost makes the problem I'm trying to solve more complex. I'm really focused on is making sure that we don't lose sight of the fact that we are at our core operators. And so we want to make sure that when we're building a solution, it's functional to the operators that are ⁓ it because we've been in that position before. speaker-0: Mm-hmm. Awesome. Yeah, I'll leave a link in the show notes speaker-1: Awesome, thanks a lot, Pete. It's been really great to be on the show today with you.