How AI Fixes Inventory, Fresh Waste & Stockouts in Convenience Retail

All right, I guess we're I guess we are already started. Hopefully everybody can see us. Um, Victor, can you see my slide? >> Yes, I can. >> All right, perfect. So, we will get started and hop right in. Um, we've got a good audience. I think people are still trickling in. Uh but we'll go ahead and get started and um really appreciate everyone's time for joining us this morning or this afternoon depending on where you are in the world. Before we um jump into the content, just want to quickly, you know, do a little housekeeping. So, uh we're going to do a webinar today. Please uh everyone's muted on the call, but if you have questions, please just enter them in the Q&A panel in the Zoom webinar screen and that will allow us to kind of view your questions and we will get to your questions [clears throat] at the end of the webinar and uh look forward to viewing and answering those. Um so without further ado, let's let's do get started. We're going to discuss sort of

the margin gap or you know gaps in this case that most fuel and convenience retailers uh really don't have visibility to and aren't able to prevent in their in their P&L. And so uh we've got a few scenarios to run through, but before we do, we'll do a little bit of introduction both about myself and Victor and also Leafio uh as a company. So I'm Brad Mitcheller. Uh I'm the vice president of North America here at Leafio and uh you know I've got a number of years of experience in retail and supply chain and I'm really focused on helping um clients both in North America and all around the world uh to leverage technology to impre increase their KPIs automation etc and retail and then Victor if you want to introduce yourself as well. Yeah. Hi, my name is Victor Hart. I'm senior account executive with Leafio and have uh experience for many many years working in supply chain specifically focused on retail operations and helping retailers

get better results, better business value out of all of their supply chain operations and looking forward to walking through some of the scenarios today with you folks and hopefully enlightening you on some opportunities to to improve your business. Yeah, Victor and I both talk with retailers around the world all day every day. I mean, this is this is the role that we have. And so, um, you know, sort of the structure of the webinar today is to talk about some of the challenges that we see specifically in convenience stores, um, retailers dealing with. And so, we'll go through that a little bit about Leafio. So, Leafio um, you know, we're a software provider. We implement solutions all around the world. We're in, you know, 40 plus countries and and have 250 plus projects, um, 15 years of experience, etc. Which really allows us to leverage our expertise and, um, you know, help clients in the best way possible improve their supply chain and merchandising process. When we think about that, we, you know, we offer this

unified platform, uh, the Leafio AI platform. And [snorts] there's a lot of modules, what we call modules here. Today in our webinar, we're specifically going to be focused on more of the inventory optimization side of the platform, but wanted to take a quick look at all the things that we do. Um, just to orient our audience for uh a little bit about our company. So, let's start in the upper right hand corner of this. We offer, you know, on the merchandising side, assortment performance, which is really answering the key question for our clients. What should we sell and where, right? It's probably the first and most critical question that a retailer has to ask themsself. And so we use uh you know datadriven AI to make decisions and recommendations based on the assortment uh at the store and cluster location for our clients um along with much more. Once we know what we're selling and where we have to make the decision as retailers, how should we visually merchandise these products in our

stores? And so that's where our shelf efficiency you know planagramming module comes into play. So it's creating automated planagrams based on you know company strategy and best practices. Um it's uh you know measuring the results of those planagrams offering suggestions etc. And then critically helping with compliance um both via our mobile app so that store managers and store employees can track the success of implementation of planagrams uh via AI image recognition. So now we've answered the question, what do we sell? How do we visually merchandise it? Then we move over to the supply chain side and now we have to make the decision. Okay, I know what I'm selling. I know how many facings and where it is in the store, etc. How do I replenish it appropriately so that our inventory in the store and in our warehouses is aligned with my demand across my stores in my entire network. So we have you know demand forecasting, automated replenishment and much much more on the inventory side as

well. We have a lot of functionality around fresh goods which we'll talk about today. Um promotions, machine learning forecasts, um [snorts] so on and so forth and uh we're going to go deeper on this side today. So you'll learn a little bit more. So that's a little bit about our platform [snorts] in terms of who we serve. You know, we serve clients in various industries. And I really like this view of our clients because we've got some great, beautiful clients that are, you know, globally recognizable names. But I think really the important takeaway here is that, um, you know, we work with all kinds of retailers. grocery, you know, convenience store where we'll focus today is one of our biggest segments, um, liquor, pet, pharma, DIY stores, and, um, as well as some CPG and distribution as well. So, you know, a very broad uh, focus in terms of industry. Um, and we're really going to focus on, you know, the gas and oil space today where we have many, many

clients using the products. All right, so that's a little bit about us. Hopefully uh everyone's still awake and engaged in our webinar. We're going to get to the good stuff now. So um you know the focus of the webinar today is where are we leaving margin behind in our business, right? You know we know based on our conversations with clients and then implementing our solutions and seeing results that the typical convenience retailer is leaving margin on the table. Um, and it's not just, you know, one place where you say, "Hey, we can flip this switch and increase your margin." What's happening is, you know, we have clients with big retail chains and they have all these little decision points that have to happen in order to run their business. And at each stage or level, you know, we can identify this margin leak and um and help to correct it using um, you know, digital transformation and software. [snorts] And so, uh, we're going to talk through that a bit today. The reason why this is occurring though is because gas, you

know, fuel and convenience retail has changed significantly so much over the last, you know, few years, several years. And I think there's a lot of trends going on in gas and retail. But sort of what we see in the market is that, you know, gas stations are so much more than just, you know, grab and go, you know, stock items, get fuel, etc. They're really becoming retail hubs where uh our customers at these stores fully expect to be able to go in, get their staples, rely on this for everyday goods, and uh and our gas stations have to evolve to really uh accommodate that need. The way they're evolving is many, but I think three of the big ways they're evolving is, you know, very critically fresh food, right? The convenience stores that we talked to are putting more and more of an emphasis on the fresh goods category, fresh foods, grab-and-go options, etc. Um, and this is uh helping drive traffic in our

stores, right? If I'm on the way to work in the morning, I've got to have breakfast. So, I pop in and, you know, fuel up, get food, buy some goods, etc. So, it's a key driver of margin traffic in our stores, but it's very complex to manage because of perishability, um, waste, etc. Uh, so that's one way. A second thing that we've seen throughout the retail chains that we talked to is an everinccreasing assortment that is just expanding over and over with the retailers that we're working with. I think retailers feel like to in order to stay competitive, you know, we have to offer more differentiated product and assortment. It's no longer enough to just carry, you know, the the staple items that we are used to seeing in in convenience stores over the years, but many of our customers are adding regional and local specific products to the range that hopefully further entice those local customers to uh you know make our store their home on the way to

work, the way to school etc. More SKUs and more assortment of course means more complexity, right? and a lower concentration of sales um you know lower runners etc which are very very difficult to manage and plan um and I think everybody on the call probably probably knows what I mean by that [snorts] lastly I think one of the the key complexities and uh transformations that we've seen in the sea store space is an increased reliance on promotions and loyalty programs etc right ways to reward our customers customers and you know earn their repeat business, gain loyalty um is great and it's a key part of the strategy for the clients that we work with but it drives a lot of complexity in terms of you know adopting these methods around promotions or loyalty um rewards but also executing them properly. And um you know it's it's it's

a really difficult complex thing to do and when we've got 50 100 200 stores etc it can become a real nightmare to manage and so um you know we talk about that a lot as well. [snorts] So, in general, I think everything's just getting a little bit harder. Um, both from a process and technology standpoint. the clients that we work with are looking for ways to [snorts] improve their KPIs, improve their efficiency and productivity, right, by automation um and ultimately make their day-to-day jobs a little bit easier and uh and up to the task of these changes that just talked about the challenges in retail that people are experiencing in sort of the new age and and the need for transformation and evolvement um in order to keep up with these challenges and continue to grow our business and uh and you know our productivity etc. So what we're going to do today is walk through a few operational scenarios like you know real things that we see our customers dealing with related to these challenges. Um you know how they're

managing it today what the challenge is the pain and then we'll Victor will kind of come in and show how we deal with the same situation inside the Leafia platform to sort of in many cases automate resolve and improve the pain uh that we're experiencing around this. So we've got a few scenarios to go through and then at the end of course we'll um we'll talk we'll talk and do some Q&A around these uh these different topics. All right. So um the first scenario we wanted to talk about is sort of you know this decentralized manual store planning that we see which I think has a really large impact ultimately on the P&L for the clients that we serve. So um you know I don't know Victor and I we talk with retailers in se store all the time and it doesn't matter if we've got you know 50 stores 100 stores even 200 plus stores uh we still run into people that say yeah you know how is ordering done just manually right our store managers they walk

around the store each week and sort of decide what to order based on what they see what their tribal knowledge is they plug it into the system and uh and sort of cross their fingers and hope it's correct. So, I think um you know the major challenge with this really is that each store manager is doing it a little bit differently, right? We've got Bob in one region who orders to protect availability. Um but Jim likes to keep a lean store, so he orders much less. So at the end, Bob has overstocks, Jim has shortages, and neither of them really can consider the complexity that comes along with, you know, market demand trends, seasonality, upcoming promotions, supplier issues, all that stuff, right? And and the outcome is really that we have a very fragmented replenishment process that results in variable store performance. Um you know overstocks in places where we don't need it, loss sales in other places where we

don't need or where we need units uh perishable waste and and a lot of manual work as well. Right? These are our store managers. We're asking that we're relying on them to do a lot of manual work in order to uh to replenish our stores. Um, so Victor, >> yeah, [clears throat] and this can have a huge impact on the P&L, right? Especially across a really large network of of stores. Um, you know, you're wasting a lot of time on manual processes to manage inventory, still suffering from that margin loss we talked about. Brad mentioned the loss sales in some situations, overstock inventory costs in others. And so what I want to do is take a look at how companies can transform from those manual fragmented processes into a standardized automated and ultimately resilient um replenishment system. So I will take over the screen share from you Brad >> and take a look at the inventory optimization solution that Leafio

provides. And so ultimately what our goal is going to be is to keep these indicators down here at the bottom as low as possible. Keep our overstocks as low as we can while preventing as much loss sales as we can. So how do we do that? Well, what our system is designed to do is autonomously uh build, forecast, and plan for demand and then ultimately build dynamic stocking buffers for items at the item location level that are going to be resilient enough to effectively manage our demand on an ongoing basis. So, it picks up on seasonal trends, promotional aspects, the perishable um items that Brad mentioned earlier, etc. And it's really critical for us to manage these at the item location level because my demand for this water at store two is going to be different than it is at store 22. Uh the lead time might be different etc. So we want to manage them independently. And then what we're

seeing here is the dynamic nature of the stocking buffer in the background of this chart. And the idea is that as our demand, as our consumption from our customers fluctuates throughout time, the inventory levels will fluctuate in response to that. And so we can see in this example, I've got a yellow line down here at the bottom. That is my consumption trend. We call it ADU or average daily usage. And you can see that as that consumption is shrinking over time, the buffer for this item is shrinking in response to that. As the demand increases, our sales are increasing, the buffer in response will start increasing so that we're always pacing inventories to what that demand is. There are several zones within this buffer. So the green zone that you can see is set to cover essentially how much inventory we're going to need over a given replenishment cycle. But we know that that's not going to be a 100% accurate uh measure. So, we need some

extra protection built into this buffer. That's where the yellow and our red zones come into play. They are the safety that's embedded within the buffer. Every single item is going to have a yellow zone and it's there to protect against volatility. That could be variable demand patterns. That could be variable lead times. If my suppliers, whether that's an internal supplier or external supplier, are inconsistent with their deliveries, I need to be protected against that. The red zone is an additional layer of protection uh that's designed for our critical A items. So items that we never ever want to stock out on, we can add in some additional protection for those. These zones are very similar in concept to a traditional safety stock. The big difference being the dynamic nature of the buffers. And then very critically, the system is automatically going to pick up on um how variable an item's demand is, which is going to influence how thick these buffers need to be. So a really steady

runner doesn't need too much protection, while a highly variable item needs to have some more protection built into the buffer. The system is automatically going to pick up on that for us and automatically adjust that throughout time. So if a steady runner becomes highly volatile, it's going to adjust that automatically. Again, picking up on those trends faster than a user might be able to just by manual processes. Lastly, within the buffer here, we see a gray zone. That's going to be the amount of product that's actually on the shelf. Um Brad mentioned our shelf efficiency solution, which builds planagrams for our clients. And so we're picking up on those quantities of facings directly from that system and building them into the buffers automatically. And then ideally what we would see over time is what we see here with this blue line which shows the actual inventory levels for this item fluctuating throughout that green zone within the buffer. Occasionally dipping down into the

yellow or the red zones. That's perfectly fine. That's what those zones are designed for. But this gives me a really clear um vision in terms of how effectively I've been able to manage my inventories over this given time period. And these buffers are calculated automatically and then when appropriate, they're going to result in order recommendations for us. And so here I can see an example of a a new supply order from an external supplier number 62. This could be from an internal supplier, my own distribution center. So we can structure the supply chain however um your supply chain is structured. And then from this supplier we see all of the inven or the uh the recommended order quantities for um given items as of today. Very critically if I want to optimize this order for whatever reason maybe I'm trying to keep it under a certain dollar amount or fill a truckload or something like that. We have optimization um functionality built

into the solution. So I can optimize around the dollar value, the volume, etc. Fill containers, fill truckloads, etc. In this case, maybe I don't want to order $18,000 from this supplier. I only want to order 15,000. And so I can input my target values, optimize, and then the system is going to update the order quantities where it can while still respecting ordering parameters like minimums, multiples, etc. Um, and then slightly adjust the order quantities for each item. Once I'm happy with this, then I can click place checked rows and then that would generate the corresponding in this case purchase order to be sent back to the ERP or the order management system uh for execution of that transaction very critically. I walked through this in kind of a manual process today, but this kind of automation or this kind of optimization and the order generation can be automated and ideally will be

automated. So, where we want to get is where the vast majority of our ordering is going to be automated, which frees up a lot of time for our users to be a bit more strategic and focus in on exceptions within their supply chain. So, in this example, I can see that I'm automating 97% of my replenishment, which is very consistent with where most of our clients get. Um when we first implement with a client, they'll start um kind of small, but over time once they mature and grow in confidence in the system, uh we have clients automating well above 90 95% of all of their replenishment. >> Yeah, thanks Victor. I think that's the critical thing like it we like to walk through the full functionality to sort of see how the systems working in the background and how these buffers are resilient and um you know liquid actually changing with our market demand. But the reality for most of our customers is when you know when we log in in the morning as planners or um purchasers etc. the planning is already

done right we we walk in it's already done. My store manager he doesn't he or she doesn't have to walk around the store and order anything. Um and critically we've standardized the process across our entire retail chain. So, you know, we're getting the same sort of output and results and improvement on KPIs across our chain without having to convince, you know, 55 store managers that we've got a new way of doing this or a new way of planning. And uh and ultimately, you know, it saves time, it improves KPIs, and it allows us to make strategic decisions around planning at a, you know, chainwide level, right? So if I make a tweak to the logic, it's going to roll throughout our whole network and uh and just have tremendous benefit there. Okay, thanks Victor. Uh I will share again and um we'll talk about the next scenario that I think is prevalent and we're seeing more and more of in the market. So um second scenario here is really margin

that's leaking from perishable products, right? I talked earlier about one of the key challenges that we see retail uh retailers facing right now is this increasing need to place fresh and ultrar products in their convenience stores you know grabband go items etc breakfast lunch in many cases um I think especially the the convenience store near my house like I'm there several times a week for different you know grabband go snacks and uh it's really hard to manage right um you know I Typically, we see people ordering more to be safe, right? Standard operating procedure. If we're not sure, we round up a little bit. And um you know, that's all well and good while those products are selling. But typically what happens is that at the end of the week or the end of the day when we've got these really ultraring perishables on the shelf that that didn't go in in the relevant time frame. So what happens, you know, that's where shrink occurs, right? we we have to

write this off. We you know in uh produce waste um and that really actually eliminates the margin that we've gained by ensuring the availability and protection on those items. Right? So the the buffer that we've built by ordering more has just really become a source of margin leak. You know it's not a delivery problem. It's not a supplier problem. It's not a demand problem. The problem is is that, you know, our replenishment system, whether manual or automated, doesn't typically account for the complexity of these um, you know, fresh items, both from a perishability standpoint and a demand variation standpoint, right, Victor? >> Yeah, that's right. And and so, like you said, the the the issue isn't the safety stock that we're building into the buffers like we talked about or anything like that. it's connecting that to the perishability of the item and so we need to consider the two different problems we have. We don't want to run out but at

[clears throat] the same time we want to prevent those write off. So it just adds that extra layer of complexity which our system is designed to help effectively manage um just like we looked at a moment ago. So if I could share my screen again. >> Sure. um whenever we're building those buffers like we looked at or building the forecast for our items, um when we add in the fresh components, it's going to add that extra layer of complexity. So here I've got a list of all of my fresh items. And so um along with the demand variation and the other ordering parameters, etc. that we have, we have to take into consideration this remaining shelf life for all of the items that are all of our fresh items we're putting into inventory. On top of that, we're going to have fluctuations in terms of our demand um not just on a weekly basis, but a daily basis and even during different periods of the day, morning versus afternoon versus evening

for some items. And so you need a system that's going to be able to take that day of the week variation into consideration and even that intraday um planning as well. So winter deliveries coming in for our fresh items that's going to impact um when we need to replenish whatever that schedule is can tie into the forecast uh of course. So, if I look at a forecast for one of those fresh items that we had there, um the system again is going to autonomously be building this forecast for us and you'll see the fluctuations of the forecast throughout time. That's going to be tied to the demand variation that we see for these items. Um and then the forecast is going to input or impact that replenishment process taking into consideration that uh day of the week variation, the intraday variation as well. Um promotional aspects might be a component of of that as well. And so it's all going to add that extra layer of

complexity that you really can't manage effectively if you're using a manual process. you need an automated process like um like we can help enable in the Leafio platform to pick up on all of those trends and make adjustments um as necessary. And then one of the things I mentioned earlier with the buffers, if I open up the buffer for this item, uh the volatility impacting those different zones within the buffer. So you can tell that this one looks a little bit different. So we've got thicker yellow zones, thicker red zones on this item than the first one that we looked at. That would be because of the volatility with this item, but it's still going to be taking into consideration the remaining shelf life for the item, taking into consideration the delivery schedule. And then crucially for our fresh items, we can set target service levels um so that we can say we want to maintain a certain level of availability for the items while minimizing our shrink down to a certain or or a certain

threshold. So we can build that kind of logic in to again the system automatically take it into consideration and build it into these buffers for the items again at the item location level. >> Yeah, very cool. I mean it's it's not about you know ordering more or less, it's about ordering smarter, right? And really taking into those you know account those additional factors. How long do we have until this perishes? um you know what is the intrae demand variation right um we can't plan these the same as we plan canned goods um because uh we really need to factor in well it sells more on a Tuesday morning than it does on a Friday evening right it's a breakfast bar whatever [snorts] um and so you know the system does a really good job of capturing that and ultimately you know reducing waste and increasing availability by really aligning inventory to market demand um where it matters Okay, cool. And I'm going to show again

for the next scenario, but just a quick reminder, please, if you have questions about these insights scenarios, you know, your business, um, put them in the chat and we'll certainly get to those at the end of the webinar. All right, so that's brilliant. Thanks, Victor. Um, one more scenario I wanted to go through here was this. This is becoming more and more of a topic of conversation with the clients that we talk to especially in the larger chains um where we've got hundreds or you know 100 plus stores 200 stores etc. It's really the the network imbalance of inventory, right? The scenario is very simple in a simplistic way. Uh you know, we've got one store out here um who's out of their top energy drink and has been out for three days, right? So the system is, you know, or their store manager is ordering and ordering, but it's not going to come for, you know, the end till the end of the week or next week, etc. I've got another store um who's overordered it and has way too much or the demand is softened there

etc. And you know, we really just don't have that visibility or connectivity between the two stores to say, "Hey, we can help here." Right? One store is losing revenue, another store has working capital and space, critically space locked up in slowmoving inventory. And the only thing our traditional system or manual process is going to say is, "Hey, this store just needs to order more and bring more into our network inventory." when the reality is there's an opportunity here to you know that that's a very smallcale example but to rebalance inventory amongst our network um in a powerful way to avoid bringing obsolete inventory more obsolete inventory into our full network. [snorts] >> Yeah. And this happens a lot especially when um networks are managed in that fragmented isolated um way. Brad mentioned Bob and Jim I think earlier. So Bob has too much inventory. Jim has too little. And so Jim might see that

and say, "Hey, I need to order more from our vendor or order more um to my store from the distribution center." When in reality, we can just rebalance across the entire network. And so our system is designed to do that really quite effectively in in an automated nature. So, uh, when we go through kind of the day in the life, the vast majority of our replenishment, like we talked about, is going to be automated. And the goal of everything we've looked at so far is to keep the right amount of inventory at the right location. Um, whether that's at the store level, the distribution center, etc. But we know that life happens, right? It's not going to be 100% correct all the time. And so we can then look for these areas, these kind of edge cases to um get even better performance. And so what I've done here is I've filtered down to look for um I'm looking at one specific location. We could look across our entire chain if we

wanted to, but it's identifying the system is identifying items where we're overstocked in this location and we might be able to redistribute to other locations that need to be replenished. And it's based on the BTB rate that I'm looking at here, which stands for balance to buffer. Essentially, are we carrying more inventory than we need relevant or relative to the buffer that we have set up for the items. And so I can find all of these items where we're overstocked. And then I've got the opportunity to redistribute these stocks across our network. I could just distribute the overstocks. I've also got the ability to, let's say we're closing out a store, I can redistribute the entire balance from that store to other stores if I wanted to. But in this case, let's just say we're redistributing all of our overstocks. And so what the system is going to do then is it will analyze the inventory levels across our entire chain and then identify opportunities for us to send our

overstocks over to other stores that need to be replenished. I can use my skew packaging if I want to. I can also redistribute back to the central warehouse. So if I have remaining inventory um after that distribution, I've got the opportunity to then send it back to the DC if I want to. And then just like we talked about earlier, crucially, I can automate this process if I want to. In this case, I'm not going to do that, but we can build that into the automation for our general replenishment process. And so then based on the analysis that the uh system is doing here, it's finding all of those items. in this case at store number two that I'm overstocked on, identifying other stores that need to be replenished. And then where possible, it's going to generate special orders for me to send from one location to another. Um, it'll take just a moment here to run all of that logic because

it's doing a lot of analysis for me. And then once it finds that, you'll see it's generating five special orders for me. Um, I'll have the opportunity once it creates those to review those. I can adjust the order quantities if I want to, just like we looked at in the order automation, I can optimize these order quantities if I want to, etc. And then once I am happy with them, I can then create these special orders, which again once I generate the orders in our system, it will send those back to the ERP system to create the corresponding transactions. And then one very critical I think piece on this is um we we need to put some guard rails in place here, right? If I'm overstocked by one case of that energy drink in one store, it probably doesn't make logistic sense from a cost standpoint for me to send it over to another store because I'm not that overstocked and it would cost too much just to send that one case. And so

we can put those kind of guard rails in place in terms of like minimum count size and the units we want to redistribute or what have you so that we're staying in our relative balance across our network without incurring unnecessary costs from a logistic standpoint to try to redistribute there. >> Yeah, definitely. I think this is something that um you know most retailers and sea stores don't really have the ability to execute. definitely not to automate, right? You know, it's it's it's hard enough for them to even manually go through and say, "Hey, what really should we be transferring from store to store or back to the warehouse, etc." And what that does is it just drives a lot more obsolete inventory in your network when you can really, you know, rebalance a bit and just save a lot of working capital on, you know, purchasing new inventory. Um so so we provide the analysis in an automated fashion but also the ability to automate it where we say hey we want to look at this report once a week and it's going

to recommend to me what I should or should not um uh execute in terms of transferring. So very very powerful has a huge impact on um the P&L and also you know our our balance sheet in terms of inventory but in terms of the P&L it's about creating availability of the products that are selling very very quickly right and and preventing those loss sales. All right let me share my screen again Victor and we will press forward. So um yeah those are sort of the three scenarios that we wanted to walk through today. I think three scenarios that we're hearing in every conversation we have is sort of top of mind for convenience store retailers. So, you know, how do we standardize and automate our store planning process to create results that are, you know, um across the chain? You know, uh planning fresh inventory to a better degree, right? really taking into th those complex factors and creating a better plan, forecast and replenishment method for fresh and then

when we have imbalances, right? Being able to execute those um to move them to different stores, different locations to prevent um obsolescence in our network. So these are three things that you know we do, we hear a lot about and we we correct. But as I said at the beginning, I think the margin loss at sea stores, it's happening over all these little micro tactical um parts of the business. And there's many many more things that are driving margin loss that the system and the platform help to overcome. And ultimately you know when we start automating these things start analyzing them we see tremendous boost in KPIs you know reduction of loss sales um you know reduction in inventory and then very critical you know productivity gains right it's a reduction in the human capital required to just stay on top of my 200 locations with 5,000 SKUs [snorts] etc. Okay. Um, so that's what's really missing is this decision layer which I

just talked about, right? How do we how do we connect these different things uh together under one platform where we have AI looking at and making decisions for our business. Um, an example I always like to put in a real life example of a client that leverages you know this technology uh and our process in order to see tremendous results. So, one of our great clients, Baltic Petroleum, you know, 86 gas station and convenience stores, 2500 SKs, so very very typical. Um, they've implemented the solution, uh, and really had tremendous results in terms of return on investment in a very quick way, right? Sales increased by 62% with, you know, availability increasing while uh, decreasing excess inventory to the tune of uh, of of half. So, pretty amazing results there. Um, 9% loss sales decrease and 8% average skew availability increase. So, it's a tremendous thing, but we're actually

decreasing our inventory and increasing our availability by aligning inventory with the right market demand and it being able to fluctuate and follow the path of demand over time. So, tremendous results there. Um, we can see or you know, Victor, why don't you explain our simulation? But I think the point is that we can show you sort of what we expect your results to be by using the platform. >> Yeah, I think it's really really important for folks that are um thinking about, you know, going through this kind of transformational process to be able to prove that it's going to deliver value to them. So for that reason, we've designed what we call a simulation. And the simulation is quite simple. We take some historical data on your items. So, we would typically choose um you know a few hundred of your your most critical items, maybe five or 10 locations, something along those lines. Then we just need a little bit of item master data like the items lead time, uhQS, that sort of thing. And then a year's

worth of sales history for those items. And then we can plug that into our system and model what the um inventory performance would have been over that time period given the actual demand you experienced at the item location level. And we can show it at the item location level as well as roll those results up to an aggregate level. And so what we see on the the slide here is an example of a simulation that we recently ran for a prospective client. And so we're measuring the results, their actuals over a given time period to what our simulation showed across four key metrics. Our loss sales, average inventory levels, service levels, and average overstocks. And so in this example, we saw that from a simulation standpoint, had they been running with our system over this time period, their loss sales would have been um x percentage lower, their inventory levels

um looks like 15% lower, service levels or availability would have been higher, and their overstocks would have been much much lower. And so there are obviously some assumptions that are going to be built into this process, but it does give us a very clear idea of the impact that our solution could make given a client's real experience, real world data. >> Yeah, it's awesome. It's super black and white, right? It's like, hey, here's how you planned last year. Here's how we would have planned. Here's the difference, right? Here's the delta. And so it becomes a very black and white conversation uh coupled with sort of the automation that the solution creates. It's often a super fun, engaging and like educational um type exercise for for our businesses that we work with. So please reach out if you're interested in, you know, sort of running one of these simulations. So that concludes the content that we have. We we've got a few more minutes. Um and we do have some questions in our Q&A here. So we're going to go through those. Thank you everyone for uh

sticking around and putting in some of these thoughtful questions. Uh and apologize again for the power outage that we experienced. I noticed Victor when you were presenting like in the bottom left of your laptop you've got like the weather widget you know and it says like wind advisory like high wind. So um you know just just leave it to Atlanta weather to give us crazy wind and down trees in the morning when we have a webinar. But uh appreciate everybody's patience there. All right. So let's go through some of these questions, Victor. Um the first question is in the chat. So what usually changes inside a retailer once these decisions become automated rather than managed through spread spreadsheets? Um Victor, why don't you take that? >> Yeah, I think well one of the first things obviously is the uh the amount of time the manual work it takes. So a lot of times users are just living um in Excel doing all of these manual calculations. um or in many cases we talked to uh seestore clients where

their store managers are just you know going they're eyeballing their inventories on the shelf and um so like Brad talked about earlier that's uh very timeconuming and just a very fragmented non-standardized way of doing things and so the first thing is the reduction of the amount of time it takes which is frees up their time to be more strategic so one of the things especially within our solution is by automating this process there's a tremendous amount of analytics and reporting available. So rather than doing those kind of manual um calculations gives you time to focus in on the analytics and uh search for the exceptions, search for the problem areas and then be strategic about um what you can do to resolve those problems. Just a quick example that I see a lot are we can track the reasons for lost sales like suppliers were late with their deliveries or simply didn't deliver the the the products that we ordered and so we can find those kinds

of issues and work to resolve them. So saving time focusing in on the real priorities instead of just that kind of manual um calculation in spreadsheets. Yeah, it's a there's so much more um I think we could say about that, the benefits that we see from people, you know, automating these processes, but you know, it's a very good overview and we'd love to discuss that more with you. Um okay, great. Next question. Um how does the platform actually connect decisions between uh Okay, so this is a question about you know, I sort of showed the platform that we offer and we focused on inventory state. questions around how does how does you know it connect between the replenishment and the supply chain side with promotions self-execution etc and um and I'll speak to this so I think one of the our huge you know messages when we talk to clients is you know we intend to connect the supply chain and merchandising process and

technology for clients that we serve and we believe there's a tremendous benefit from doing this typically clients that we talk who they have disconnected, you know, teams and and technology and process for these two sides of the business. And the reality is is that it becomes a really cumbersome manual process and challenging to actually execute the decisions that merchandising makes um because they're not connected. So if I change for instance um I make a decision and make a change to the assortment, right? How does that flow to the store in order for the store to update the shelf with the new assortment, right? How do we measure compliance in the store there? Also, how does that flow into our replenishment process where I've got to turn off the replenishment at those stores where I'm changing assortment for the old item, add the new item, update the central warehouse values in some cases, etc. It becomes a really really cha uh big challenge for companies to deal with. And by connecting these platform elements, every decision that I make on

one side or the other flows to the other side of the business and um it really drives a lot of automation and and synergy between uh different elements of our supply chain merchandising process. At the same time, you know, we need to know um how many facings we have on the shelf which is connected in our planagram solution to plan inventory properly. So we always have sort of that minimum shelf stock of the number of facings. Um so again another another one of many use cases where the platform becomes very very powerful. Okay. Um another question Victor maybe uh time for a couple more. In projects like this where would we see the fastest financial impact waste reduction availability improvement or labor savings? >> [snorts] >> I would say that really depends on where you are today. [laughter] Um, we can see uh improvements in all of those, but you know, if you're um a lot

of times clients will be really overstock, meaning they're not really suffering from availability issues, but they're tying up capital and product that's not moving. That ties into the waste for perishable items. And so we would typically see in that kind of environment reduction of those overstocks, reduction of that waste, etc. while maintaining high availability. Whereas sometimes clients are in the exact opposite um scenario where they don't have the the correct demand signals. They don't really know what they need to have on the shelf. So they're suffering with availability issues. So in that instance, we might see inventories go up, but that makes sense because then that brings their loss sales way down. Um, so where companies typically see the fastest results really just depends on where they are today. Where do they need to see the improvements? The labor savings for sure. The the automation we typically see in any scenario a lot of a

lot of reduction in the amount of time it takes to go through this process. Yeah. Uh totally agree. It really depends of course as does everything, but um you know I think uh benefits across across all scenarios there. Um all right, last question we'll have time for today. Uh how does the system handle very short shelf life products like fresh food prepared in the store or you know stuff with daily deliveries? Can the forecast adapt fast enough? Certainly can. Um, you know, I think this is one of the really powerful things about the solution when we have fresh goods, um, is that we have to we have to adapt to it fast enough in order to prevent those stockouts while protecting availability. Um, you know, in the solution, we actually are considering for daily deliveries, like you mentioned, we're even considering the time of day in which those deliveries occur, right? Because when you have an ultrarresh product that's highly perishable, it matters whether it is received in the morning, in the afternoon, or overnight, right? And so those kinds of things are it's data that

we use in the system to make sure we're planning um uh for those goods appropriately and replenishing those goods appropriately. >> [snorts] >> uh and certainly in terms of the forecast and the buffer logic we we are you know adapting fast enough because the forecast the buffer the weekly variation ratios that Victor talked about you know they're constantly evolving with changes in demand right so no no two days are the same and and you know our algorithms and forecasts are constantly working to catch up to align themselves to market demand and and it can adapt fast enough these items obviously have very very short lead times in almost all scenarios. And because of those short lead times, it means that we can adjust the um the logic or that the logic will adjust itself and replenish itself to uh to to adapt fast enough um just via the short lead times that we have. Okay. Um cool. Well, thanks everybody. I think there's a few more questions, but

we can respond to those sort of uh via email with our team after the webinar. Thanks so much. We've got a great audience and uh you know number of folks here. So we very very much appreciate your time. Please reach out to us with questions or if you'd like to see a simulation of your business's data. We're certainly happy to uh facilitate that. Apologize one more time for our power outage in our brief uh intermission there in the webinar. But thanks everybody for sticking around and uh thanks Victor for the presentation. I think it was a lot of fun and we will see you guys next time. Thanks so much everyone. Hope to talk to you soon. >> Cheers. [snorts]