Sales forecast automation for planning

How to create outlook signature

well good morning good afternoon good evening wherever you are in the world welcome to streamline's AI supply chain planning webinar series my name is Keith Drake I'm the Vice President of Sales North America for streamline our topic today is best practices for demand forecasting and inventory planning in 2023 a very relevant topic Our Guest today is also from streamline Malcolm O'Brien he is a supply chain expert here at streamline hello everyone uh pleasure to be here and to meet all of you really looking forward to seeing your questions and having a productive conversation thank you Malcolm I'm looking forward to it as well we can go ahead to the next slide we'll discuss the webinar Logistics today in your Zoom console there's a q a panel please submit your q a session questions at any time in that window and at the end of the webinar we've we've carved out plenty of time to address your specific questions both relative to our best practices that we'll review today and to any topic and supply chain planning um in the chat window right now we're posting a link the link will provide a handout to streamla about streamline and it's available for you to click on and download right now we will perform poll during the webinar to learn more about our audience and what your specific supply chain uh planning challenges are and then we always get this question a recording of This webinar will be provided in a follow-up email to all registrants so let's go to the next slide and we'll overview our presentation for today Malcolm if we go to that there we go all right a little bit of a little bit of a time lag we're we're located all over the globe here so we're subject to that thank you uh just to overview streamline the company and our platform again we're very excited to be here today um our topic is very relevant to the supply chain industry and I think we're going to have a very valuable discussion for you uh just to briefly comment on on streamline independent sources such as G2 recognize our leadership in providing supply chain planning Solutions since 2009 we've been around for a while our footprint is global with over 200 implementation partners and more than 1200 clients throughout the world our customers span all Industries and all components of supply chain from manufacturers to wholesalers to Distributors and retailers our streamlined solution is an AI based platform for demand forecasting and inventory planning and out of the box our streamlined solution provides very accurate demand forecasts and robust inventory replenishment plans based on your historical data your current inventory your orders in transit and any ordering constraints that you're subject to as we will see today many areas of the streamlined platform can be tailored to meet your business model and Industry conditions and your specific business processes we're going to try hard not to make this a commercial of course but we will use the streamlined platform to illustrate the best practices that Malcolm and I will discuss today before we get started let's take a brief poll if you could post in the chat window where you're from we just like to get a bit of a handle on where everybody is located um I see Mississauga Canada I've I've been there South Africa we've got two continents covered Knoxville Tennessee India uh Ontario Canada San Diego California hello Andy uh wow all over the place Cary North Carolina Mumbai India Guatemala Tampa Florida Londonderry New Hampshire truly all over the globe I probably missed the vast majority of our attendees I apologize for that but very exciting that our topic today is a is one of global interest all right let's um let's tee things up and kick things off today our webinar today we'll discuss some common supply chain planning challenges such as supplier unpredictability which I'm sure we're all subject to historical sales data and distribution data that are impacted by various disruptions such as as were some of us are still experiencing the global covet pandemic and then always a challenge forecasting the demand for new products whether they be totally new a replacement product a substitution product we'll be reviewing some best practices for that as well before we dive into our first topic though I want to share some uh some prefacing thoughts you know supply chain unpredictability is the new normal for sure it has been for at least a couple of years and it will be for the foreseeable future our jobs our responsibilities are very challenging I read recently in a world economic Forum report that senior Executives in operations and in Supply Chain management that they expect the impact on disruption on corporate value to increase by up to 25 percent over the next few years and that only 12 percent of companies are adequately protected against future disruptions in supply chain and operations so I think that tells us that many of us are aware of the issue we react to various uh issues but we're not prepared to proact some of our best practices today will hopefully shift your focus from reacting to being proactive in response to these uh these uh factoids that I just shared with you Gartner reports that 23 percent of supply chain leaders expect to have a digital supply chain ecosystem by 2025. I know many of the folks that we talk to who are interested in our platform they start the conversation with we are transitioning to a digital stack for all of our supply chain planning and management uh so it's good to see that shift in Focus but I think across the industry it's still ongoing so each of our three topics today deal with uncertainty and how to squeeze maximum risk out of your supply chain equation in the specific areas of demand forecasting and inventory planning as I mentioned previously after we suggest best practices in these three areas we'll briefly show how to implement the approach and streamline with a quick demonstration there are certainly other ways to implement what we're suggesting as best practices I would suggest to everybody that the key focus is does the best practice make sense for you your company your business processes and more importantly how can you automate the implementation of that best practice uh Malcolm anything else to add before we Dive In um yeah thanks Keith I I would say supply chain unpredictability has ironically become quite predictable uh in The New Normal so it was really interesting and and frankly shocking to hear the size of those numbers and the impacts having on businesses uh I think covet has made it painfully clear that excellent Supply Chain management is not just the cost of doing business to be minimized but an opportunity right a potential source of competitive advantage and it's about preparation when the inevitable next disruption comes whatever it may be if you are the business that is more resilient better prepared then you will be the business left with improve customer satisfaction Revenue market share however you measure success hey you remind me a colleague at one of our partner organizations recently said I am interrupt driven in other words totally reactive as I commented a minute ago many of the best practices with a little bit of effort can become can transition from you in a state of of reaction being interrupt driven to being proactive and even mitigate that risk before it occurs exactly okay good comments thank you Malcolm let's Dive In our first of three best practices we're going to talk about supplier unpredictability arguably the uh one of the biggest of the of the the biggest of the three um some common examples are a supplier changes their delivery date supplier changes their order quantity or both there are certainly multiple sources of unpredictability at every point in the supply chain these common examples are just a few uh so among the most frequent and with the highest impact is unpredictably unpredictability among suppliers from manufacturers to Upstream Distributors any organization that's supplying us with inventory so we're going to take a quick look at these issues and then suggest both some tactical methods to react and strategic methodologies to reduce this risk with the common examples on the left here as examples we're going to suggest some tactical best practices reactive first and then move on to strategic methodologies uh some tactical best practices to be reactive to what seems to be constant supplier and predictability is to change the order status in the Erp to maintain a single source of Truth many of our clients have a digital technology stack of several systems it works quite well with them but the best practice is to make that change in that order date that delivery quantity Upstream so that all the downstream systems uh are singing off the same song sheet and so the Erp would then update the subsequent supply chain planning platform many planning Solutions such as streamline allow changes to specific Fields specific pieces of information like supplier lead time shipment qualities supplier lead time variance and other ordering and delivering parameters I want to talk about that third bullet briefly A supplier lead time variance if you are collecting data from your suppliers in the variance uh the the difference between their their contractual obligation and when the actual deliver that can potentially be very valuable information to do to perform your supply chain uh your inventory planning with knowing what the variances of your suppliers how predictable they are can become a very useful Factor so the summary on this slide is it's best to make changes Upstream at the source now let's move forward and talk about some strategic best practices they include and we're going to demonstrate this here in our solution platform in a second but synchronize orders of all items for each supplier let yours what that does is that lets your suppliers understand your supply needs your ordering requirements and you become predictable to your suppliers the more predictable you can become to your suppliers the less unpredictable they become with you sounds simple and there actually is a very simple um actual piece of actual information that we'll show you here in a second within streamline to affect that and what that is on the fourth bullet is the transition from a min max or replenishment Point ordering strategy to a periodic ordering strategy in other words instead of having a set point that includes Safety stock that you draw down inventory to and then reorder right then and there at that point save that and order on a more periodic basis this does reduce uncertainty it minimizes risk it may mean carrying a little bit more inventory a little bit sooner so there is that carrying cost but it will help solve quite a few of the issues that are associated with supplier unpredictability Malcolm before I grab the screen and move to the quick demonstration of streamline anything to add uh thanks Keith uh I would say yeah when when dealing with uncertainty uh Readiness and foresight enables speed which are all keys to success and synchronizing orders with your suppliers and establishing predictability or ways you can take what is out of your control you cannot control your suppliers supply chain but you can make yourself more predictable to them you can bring that into your locus of control so that when the emergency or unforeseen disruption comes whatever it is a new covid or a new the sus Canal gets blocked again both you and your suppliers have more time and attention as resources to funnel towards exception handling rather than trying to run your business and fight new Fires at the same time that's right okay good comments I'm gonna switch I'm going to grab the screen from you Malcolm if that's okay and we're going to explore very briefly in our streamline solution platform how to make yourself more predictable to suppliers a brief overview of streamline streamline has a number of major modules they include demand forecasting inventory planning a very broad range of reports both projections and historical and then a kpi dashboard both at the top level and the detailed level to measure your progress in reducing Overstock stock outs and a number of other kpis what I'm going to demonstrate now how do we become more predictable to our suppliers many systems allow the ability to define a replenishment planning a replenishment strategy rather either periodic or min max replenishment point the more desirable to become predictable to your suppliers would be periodic but I'll select min max and now what we have is um a set of planned orders based on that min max strategy what we want to see however is over time how will that work out over time will a min max strategy versus a periodic strategy actually reduce the number of orders for the same amount of inventory for the same skus from our suppliers streamline includes a very useful feature we call it time machine it's a dynamic simulation where you can play forward in time a replenishment plan against ordering constraints underlying a demand forecast to see how things happen in the future all run time machine in this min max replenishment strategy forward six months what streamline is doing right now for this very simple inventory is every event that changes the inventory every purchase order every receipt of shipment a transfer from one Warehouse to another is being calculated recalculated and stored to compute a number of things now I ran this offline earlier and here we see the results I've exported the results the result of running a min max replenishment strategy versus the results of a periodic replenishment strategy over the same period of time over the same six month period the results are with a min max strategy we are placing 84 separate orders to our suppliers over that six month period to meet demand maintain Safety stock but not overshoot and have Overstock issues if we switch to a periodic ordering strategy you can see the to order the same inventory is only 51 orders that's about 40 percent fewer orders yeah about 40 percent fewer orders just by switching to a periodic strategy now that's the simple part the simple part is switching that strategy does your supply chain planning platform support that that's up to you to answer the major challenge though this is another best practice is within your organization to shift Focus to shift policy from a min max ordering strategy to a periodic strategy sometimes that's straightforward larger organizations it becomes more of a challenge uh before we move on Malcolm anything else to add to the demonstration here um yeah I I just like to highlight some of what you said there Keith I I find that many conversation in many conversations I have many many people undervalue this switch between periodic ordering and min max um once you find the right I think Keith demonstrated it very well the number of orders was nearly half right and so through automation or or software or however you you guys decide to enable this right it's elegant in that you're doing you're ordering less so you're doing less work and there's less potential points of failure in the supply chain less boats less crates less planes whatever however you guys do business but you're also lowering costs in the Supply Chain by keeping less safety stock and decreasing uncertainty and the leaders are always telling us oh do more with less and we think to ourselves what does that mean this is one of the ways that you can do that it's about all about automation that's right decide the policy and then have a platform to automate it all right we'll move on to our second of three best practices and that's historical data disruptions um we're all subject to it the list of causes on the left that I'll review briefly here in a second are just a subset of what we hear on a daily basis from our customers and that we help them with some causes of data this historical data disruptions historical data in terms primarily of our historical sales and or distribution data so inflation and high interest rates are affecting us all causing uncertainty in our supply chain uh the Russian invasion of Ukraine that occurred a little over a year ago though many of us may say well how did that affect supply chain that had a huge effect throughout Europe on the supply chain for agriculture products uh Ukraine in many ways is a breadbasket of at least Eastern Europe if not all of Europe in many categories and the Russian invasion disrupted that supply chain and caused data disruptions uh where we're always unfortunately have global trade conflicts that we have to deal with um stockouts during unanticipated demand surges good problem to have but causes historical data disruption supplier unpredictability as we talked about a minute ago and what I mean by historical data disruptions are both artificially zero sales or artificially low sales historical data that we collected but may not be totally reflective of our industry our company and therefore shouldn't totally be relied on to make demand forecast moving forward so best practices uh is to revise your demand forecasting strategy to recognize disrupted historical data at a minimum manually as I'll show you here in a second within our streamline solution or in a more automated fashion if possible here though we're recommending that the best practice do not change your Source data in other words Upstream in your Erp The Source data or the truth data has not changed but we need to revise it within our supply chain planning environment to best take advantage of the fact that we know there's been disruptions and how to deal with them uh Malcolm I'll ask you again anything else to add at this point uh well said Keith I think I would say how do you fight on predictability you fight it with good science good data um speed I think adaptability and measurability are key here you need to recognize the change in the market create a model you think uh represents the the new market right and measure its performance going forward um to determine whether you scrap the model and start over and digitization is enables all of that right a basic statistical framework to start with a visual sandbox to test your model with historical data analytics to track performance automation to save your your time and energy and and give you more time to focus on the exception handling later makes sense makes a lot of sense but again easily said a little bit more challenging to implement at times but we know we need to do it all right I am I'm sharing my screen again Malcolm very good so here's streamline again and now we're going to focus on the demand forecasting area as a brief tour of the demand forecasting area in the upper left we see our item Tree in this very simple example we don't have locations streamline is multi-echelon regarding locations distribution centers warehouses sales region however your business operates we can replicate within the platform and distribution channels as well customers can also be represented in the item tree here in the middle upper middle we see our sales and operation planning Grid in this example I should have commented earlier the current date today's date is January 1st 2023. so historical information to the left projections to the right and then down below we see a graph of some of the data in the snop area in Gray is the historical sales a plot of this first row in blue and in red is the model that is selected and configured through streamlines AI based processing and blue is how well that model fits the historical data and then in red is the demand forecast moving forward and let's see what have I picked here in this case for this SKU streamline has suggested that their seasonality and Trend information in the model but we can see here that in November and December we had zero sales let's say we actually did record zero sales but due to some anomaly and we know that our supply chain planning platform specifically demand forecasting solution should not consider that they were artificial they won't occur again we mitigated the risk um the best practice would be to Simply ignore those points now I'll Focus your attention on the graphical area down below as I click these check boxes which read ignore actual sales you can see that everything changes and now when we ignore both November and December we see a much flatter demand forecast moving forward these two zero historical sales months artificially created a trend that the platform recognized so we recognize that's not correct data let's just ignore those two months and now we have what is likely a much more accurate demand forecast moving forward uh Malcolm I have mentioned several times automation's the key you don't want to be clicking on individual skus all day that does not free up your time in Your solution implementing this practice it would be very useful to have a global setting to ignore zero sales and when we do we've we've ignored zero sales across our entire item tree and streamline has readjusted ingly so quick demonstration uh but one example of how to recognize and account for historical data disruptions Malcolm anything to add uh yeah briefly Yes actually um I would say it's important your data deal with or excuse me it's important that your your strategy for processing that data deal with disruptions uh like zero data or missing data like Keith has shown a classic is uh Chinese New Year but um also artificially High outliers so things like uh one-time promotions for example that will not occur again uh the idea here is is when your data is not representative and that's a judgment call um then your your strategy should automatically detect these situations and enable you to focus on the exception management you mentioned Chinese New Year well that's not supplier unpredictability for many of us the factories in China shut down for what a couple of weeks we know that it's coming but how do we deal with it's anomalous in terms of it only occurs one month out of the year so a best practice there is to account for that in an automated fashion as you just suggested all right good discussion good discussion so far I see there's a lot of questions coming in and and we will get to them here in just a few minutes during our q a session uh our third topic is new product demand forecasting uh just about everybody has an opportunity uh has the need rather the requirement to forecast the demand for products for which there is no sales history no distribution history some simple use cases are product substitutions we introduce model 1189x to replace model 1189 a product Replacements we have a a we replace that model 1189x by Model 1200 same general product but new features new functionality it's an upgrade versus an update as in the first volume and then probably the most challenging example is a completely new products for which we have no similar sales no sales history for um similar products in this in these cases the best practice is to model the demand so that we can forecast it using patterns or models from similar items with representative I.E similarly characteristic sales history those could be individual planning items uh where a planning item is a SKU location Channel combination maybe the same product the same SKU number but depending on the sales region and the distribution Channel its historical sales may be very very different so that's a key point to consider as well or as we'll see here in a minute entire product categories I'm going to grab the screen again but anything else to add at this point uh I would say new product forecasting is uh a balance of Art and Science right you need to get the data from somewhere it's certainly better than starting with nothing or just guessing but um you know it can be from other models it can be from the market or survey as it can be from your stakeholders um but then continuing to update and iterate that is the scientific piece of that process right borrowing data is great but just like he said if you put the same coffee in a different country right the Americans have very different coffee drinking habits than let's say the Italians so it may not be representative but when when new data comes in have a look at have a look again at your model and say is is it working as intended look at your kpis Automation and AI work wonders here for saving time and energy you know I I I often um ask folks do you have new products that you have no sales history or similar sales history for almost always the answer is yes we're looking to grow we just acquired a company we do have new products that fit within our portfolio but they fill a gap we have no sales history what do we do I'm going to show two very quick examples of how we would borrow the pattern from one SKU or product category and apply it to a new product with historical with zero historical sales here in our item tree we see we have three building materials concrete blocks nails and dimensional Lumber dimensional Lumber is the new product that we are just going to begin selling for nails we see as we saw before in the demonstration seasonality and Trend information dimensional Lumber though is not quite as interesting we see Zero historical sales it's a new product we see a thousand two by fours inventory on hand and that's it that's all the information we have down in the graphical area The Gray Line zero historical sales so despite the AI power under the hood of streamline it's best uh guess if you will its best estimate is you've not sold any in the past you're not going to sell in any future we know that's not true because it's a new product what we will do in this case though is we recognize because of our our expertise in our in our industry vertical that we're very confident that the seasonality exhibited by Nails will be the same as for two by fours so a lot a lot of platforms out there a lot of solutions a lot of methodologies to implement what I'm about to show you some are more manual what I'll show you here is highly automated we can go to the model parameters for nails and we can save that pattern I will give it a name seasonality pattern nails and when I save it that seasonality pattern that pattern that streamlined detected four nails from the historical data by applying a a toolkit of advanced statistical techniques and an AI based expert system is now available for application throughout the item tree I'll show you how to do that we're we're now looking at dimensional Lumber and here is that seasonality pattern that I just saved when I apply it we still have zero historical sales but now we have a demand forecast moving forward I should have mentioned earlier the green curve is your beginning of period and end of period inventory projections over time your system should use a combination of the initial assumption as I just showed you here but then also factor in month over month period over period the actual sales and fine tune that starting point that model that it started with to provide even more accurate demand forecasts I'll show you another example that is a best practice that any supply chain planning solution should provide and that's in the category of apparel we have been selling swimwear for quite a while we see the historical sales we see the demand pattern but we just started carrying t-shirts Beatles t-shirts four sizes zero historical sales uh previously I saved this pattern for swimwear and now what I will do is I will apply it at the category level for t-shirts actually I should have there we go so there's sorry about that there's our starting point zero sales history zero demand moving forward now when I apply that seasonality pattern that I previously saved from swimwear we're assuming that when more swimwear are sold more t-shirts are sold summer months when I select that we now have a demand forecast not only at the category level but for each of our four individual skus again best practice is to use your judgment expertise but have an automated process for taking the characteristics from a product that's going to act similar and apply those to your new product to create a demand forecast moving forward Malcolm anything uh what did I miss anything I left out here uh I would I would say I'll keep it brief just because I see so many questions in the chat but uh I think that was a great example of using existing demand patterns to for new products I would say uh it's also important to have a strategy for substitutions I think that was actually one of the questions and product Replacements so um streamline handles that automatically but uh the having a strategy that can understand that Nuance is also important because it does affect the accuracy of the forecast yeah that's a good point you mentioned product Replacements there is a feature within streamline where we can make that definition at this point in the future product 1189 is going to be replaced by product delay 11 to 89 x or it can be a substitution it can be a hard replacement or a substitution where either or can be sold streamline takes all that into account all right uh those are our three our three uh topics for today uh what I like to do is suggest you know for your option and consideration what your next steps should be actual information um we can't you can see us we can't see you but I imagine there's a lot of heads nodding around the globe right now that the uh the challenges that we've outlined apply to you probably on a daily basis for many of you and that some of our best practices would add value to your organization so I'll suggest three next steps um first is to develop a strategy that automates a response to data disruptions everybody experiences data disruptions but they're all different uh over the course of a month even I talked to several dozen companies uh their problems their challenges are all different so everybody's data disruption issues are likely to be likely to vary somewhat so think about what makes sense for you what makes sense in the context of an automated process that allows you to apply a strategy to deal with data disruption issues in a similar fashion number two uh your next step would be to develop and and launch a plan to quickly respond to supplier unpredictability we suggested one methodology that uh when our clients implement it they they almost immediately within two or three months see the value and switching from a replenishment point min max strategy to a periodic strategy but that may not work for you so we suggest to think about for supplier and predictability what works best for you how can you make yourself more predictable to your suppliers and then number three if if there's interest here uh contact us for a free consultation we'd like to learn what your specific challenges and needs are and work together and have a conversation about how streamline can add value to your business our contact information will be provided on a couple of the subsequent slides we're going to move to our q a session here in a second but before we do we're going to launch our first poll so I'll ask our producer to launch the poll and on your screens right now you should be seeing a dialogue there we go right there poll number one so we've covered a lot of ground here today in the context of three best practices but I think we've talked about all these challenges if you could select one or two of your challenges uh we'd appreciate that are your challenges producing accurate demand forecasts or perhaps automating the process of replenishment or ordering plans third challenge would be responding to supplier unpredictability number four compensating or mitigating historical data disruptions some of you probably experience more some less and then number five forecasting the demand for new products wow I see lots and lots of responses coming in I'll give everybody just another few seconds another 10 15 seconds or so to respond so if you could just select one or two uh what is your greatest challenge right now today and for the near-term future accurate demand forecasts automated replenishment order plans producing them uh responding to supplier unpredictability compensating for historical data disruptions or forecasting the demand for new products and please do answer the poll everyone because this informs our conversations with people like you have informed this webinar and this is your chance to to maybe see something on the list that you'd like us to do a deeper dive into later absolutely all right I'm going to close the poll in three two one and there we go all right I am not able to display this so I'll just briefly review verbally for everybody uh number one by far and away is accurate demand forecasts certainly and that makes a lot of sense if we can forecast what do we think is going to happen in an accurate uh fashion then we'll know how to respond to it demand forecasting what's going to happen inventory planning what do we need to do about it so accurate demand forecast was selected by a little over two-thirds of our participants number two uh forecasting the demand for new products uh almost half of you selected that uh quite a few of the folks we talked to that is an issue but that's a little bit higher than I would have expected so uh that that's interesting and then almost equal portions are answers two three and four with number three responding to supplier unpredictability uh being the most I think I'm sharing the results now I just found a new button in the zoom console so there you go there's your results right there all right I'll go ahead and kill that so we can move on to the next slide we'll make a couple announcements uh we are very active within the supply chain planning industry part of our activity is our relationships with Forbes specifically forbes.com um Alex that you see listed here he is a Forbes council member on the Forbes Technology Council and that allows him the opportunity to provide thought leadership articles every two or three months so last year 2022 we published these three articles and then in 2023 we published two more in your chat right now I'll ask our producer to provide the link to the button you see there all authors articles if you could scroll for it there we go so you'll see the button there all authors articles now in the chat should be appearing a link that'll take you to Forbes and all five of the articles that you see here okay let's move on to our q a session this is going to be very Dynamic here as I mentioned at the beginning of the webinar please submit your questions in the Q a window in the zoom console a recording of our webinar including the Q a session will be available via email I'm going to open and ask you to do the same thing all right I'm just going to start at the top here what is the best way to account for shelf life with perishable products great question perishable products not only include agricultural items of course that do have a shelf life uh that could be months to years could be as few as days to hours uh but they can be other types of products as well Pharmaceuticals and there's probably a number of other categories I'm not thinking of right now so what is the best way to account for shelf life your inventory planning your ordering system should have a methodology to recognize that and perhaps order a little bit product uh that's uh sooner to meet a future demand it's also important for your supply chain planning your inventory planning your ordering system to recognize it may be the same planning item that skew location Channel combination but a different manufacturing one A different block which each has its own date so the item tree that I showed you in streamline earlier certainly displays graphically those first three skew number location Channel but a very important fourth item that um that diff that distinguishes or defines that planning item is the expiration date okay um I'm not able to mark that as answered so I'll move on here is streamline a plug-in to Erp systems to better streamline demand forecasting will a complement sap Oracle systems uh to answer the second question well to answer both at once we have uh quite a few Integrations to existing erps couldn't call it a plug-in it's more of a direct uh connection to an API interface if your Erp is not on our list that is almost never an issue we do all of our connections Integrations in-house we have a very skilled team of developers that haven't met a challenge yet that's for sure all right so we can mark that as answered I'm going to move that just to just to expand on what Keith said we have for example some pre-existing connectors to sap different models of sap and uh for example netsuite so these are Partnerships we commonly have absolutely okay I'm scanning the questions here how to streamline deal with phasing out products at the end of life or end of contracts Malcolm is that one you want to grab because we we touched on it briefly you touched on it briefly during your rap on demo three sure um so streamline has a product substitution uh feature specifically for this and it can understand the Nuance between um a full replacement or phasing in and out or some level of uh or being sold in parallel so all of these things can be uh understood and accommodated for by streamline yeah and I'll as a best practice whether it's streamline another platform a manual methodology whatever your methodology is for taking into account end of life phasing out products set that up as soon as you know when that phase out date is the the longer lead time and I know this from streamline the longer lead time streamline has the better it can plan for that phase out and that phase out should be either a direct product replacement product a product B on a certain date in the future no longer sell product a only sell product b or the alternative sell both products until my inventory of a is depleted those are two different situations each of what one or the other is going to be appropriate to your business process but they could result in very different ordering plans uh another question for you Malcolm what is the best way to address forecast accuracy I like that uh uh thank you Keith I'd love to hear an elaboration from the audience member about what specifically they're asking about but uh when it comes to for example using forecast accuracy I I would say it's the classic scientific process so uh measure change something and then measure again so comparing forecasts to you know plant versus actual is a report that for example streamline uses very often looking at it month after month after month are we getting the forecast right what is the gap between what we predicted versus actual is it trending in the right direction you'd hope that would get smaller over time you know there are many statistical measures you can use to improve the accuracy of your prediction removing anomalies and outliers as we discussed looking at um mean absolute percentage error is a classic however you guys do that so measure it change something measure it again and be diligent about doing that that's right uh here's another question that just came in a few minutes ago I'll toss it to you Malcolm can streamline be used in forecasting and planning for e-commerce channels uh can it be used for 500 or more skus sure great question and simply yes and can it be used for 500 plus cues I would say you know the Streamline has been you know actual client use we have a client that is uh a significant retailer in the Middle East they have something like I think 250 Mega stores in different locations all with many many different products so I think it adds up to more than a million skus streamline has been you know battle tested for that and then stress tested for 10 million skus so 10 times that volume we don't expect to run into particular problems e-commerce uh is a very common business model that we work with as well so simply yes Laura asks how to streamline incorporate AI at the top level uh streamline employs a two-step process when looking at the Historical sales data stream to develop a demand forecast model the first step is a very thorough statistical analysis to identify outliers we saw two outliers today the two November and December 2021 those two periods of zero historical sales so a lot of the statistical analysis identifies outliers does something which is called identifying the relevant history depth relevant history depth is how far back in time can I go such that my data of historical sales has the same characteristics and prior to which the characteristics are different and I should ignore that data that relevant history depth is likely different for every planning item after that statistical characterization streamline then employs its ai-based expert system to take those characteristics select and configure a model type so in a nutshell that's how that works Malcolm back over to you uh how can I deal with constantly changing lead times from my suppliers so there are a couple methods it depends on how manual your process is but for example Keith brought up a good one during the webinar which is lead time variants so if you're aware of um you know some suppliers are more reliable than others and having a safety stock that is correlated with how reliable your suppliers is is a good way to save yourself headache and money in the future so that's a classic one um so another is simply sandboxing and changing lead times to simulate what that might look like so for example if you can see A disruption coming maybe you're not aware of the exact impact it will have on your business you can double your lead times you can triple your lead times run a simulation and say Here's what my supply chain would look like if my lead times for this group of suppliers were to double based on maybe based on geography and then you can look at the financial impact of of that uh unpredictability and say is that acceptable can I simply absorb that or do I have to develop a risk management plan because this outcome is unacceptable that makes sense that makes sense uh you mentioned it's a good segue you mentioned Safety stock uh in your in your comments just now Malcolm another question is what method do you recommend to calculate Safety stock what approach so that that's a great question and I would say it depends as Keith said on your your specific business circumstance and objectives um there are Classics like Keith brought up which are min max something based on service level Something based on a more statistical or fluid demand forecast um and I would say uh just ensuring that there's a pro con for all of those obviously min max is very simple right it's easy to begin you just pick a number um and of course on the other end it may not accurately reflect reality so I would say marrying your business circumstance with your your the amount of time and energy you're willing to put into this strategy and maintain it along with the level of results you expect right how how much lift are you expecting to get by changing your safety stock strategy or good ways to help you pick indeed we've got uh time for a few more questions here's a good one from Elijah how can I encourage change in a family business with the large age Gap in employees in other words how can I Implement new processes that require skill sets outside of current employee knowledge a best practice is to if if you are I'm assuming you're transitioning or at least have the objective to transition to a digital technology stack away from manual Excel based Excel driven processes if that's the case the best practice is to focus on a solution that employs a lot of Technology but is very two things very user friendly easy to navigate and provides visibility throughout the supply chain planning process um we'd like to think streamline checks both those boxes but regardless a solution that does check both those boxes are three boxes a lot of power and Technology under the hood ease of use and visibility to all users I think you'll find that that um that lack of required skill sets is probably not as great as you may think that it is uh it's something we'd really like to chat you about Elijah you see the contact information here on the screen give us a shout and we'll have a conversation and just to add uh piggyback on what Keith said there's also you know asking businesses about the services they provide around training and enablement is a good ask as well um many businesses will hand you the keys to a car and say ah there you go it's up to you and no news is good news right so you don't want to be left holding that and having to deal with it or having to build a shed from a set of tools you haven't you don't yet know how to use that's right Malcolm I'll toss this one to you this one came in during registration when when can we expect the market to be normal recovering after covid and supply shortages raw materials and lost vendors well uh if I knew the answer to that Keith then I'd be uh playing the stock market not doing supply chain but um and when I find my crystal ball I'll I'll let you know but I would say there's no such thing right there there will always be disruptions there's always something new going on and you know what was it what what year is it 2020 so 100 to 300 years ago we were talking about building trains right and then 170 years after that we went to space we went to the Moon and then 50 years from then no not even 30 years we had the internet and now we have ai right so the the pace of change is rapidly accelerating and that may not be what you want to hear but uh that that means that being able building a supply chain that is resilient and adaptable as it is able to handle the shock of unpredictability is the way of moving forward there isn't a perfect solution because the the best supply chain of the future is something we can't yet imagine makes sense where we I think we've got time and I apologize we have not gotten to nearly all the questions that came in during registration nor in the chat window here uh but I'll I'll answer Fernando's question I'm going to change the question slightly Fernando's question is how can streamline be applied to a trading company that has alternatives for the same product and these Alternatives can influence the amount to buy between them I'll change a question slightly can streamline be applied the answer is yes through streamlines replacement and substitution rules I think what you describe can be easily configured within streamline to give you that flexibility but Fernando give us a call you see the contact information love to hear what your uh what your needs and requirements are in a little bit more detail unfortunately that is all the time that we have I do want to end on time in respect of other meetings that everybody may have uh sorry that we did not get to everyone's questions there are several dozen that we didn't get to but please contact us send us a note give us a call we'd love to talk to you my role Malcolm's role a number of others on our team is to work with you to determine if streamline provides value to your organization uh if the answer is no we part as friends if the answer is yes we form a partnership so thanks uh thanks Malcolm for an engaging discussion thanks to our audience for attending our webinar today and you're very insightful questions uh you see here on the screen uh a brief overview of streamline as a reminder we stand ready here at streamline to help you with your supply chain planning needs our our streamline uh solution platform provides an AI based and we feel very user friendly and cost effective solution for demand forecasting and inventory planning so we're going to end our webinar today with one more poll and that's simply your honest feedback how do you think we did I'll ask our producer to flash that poll up now five choices excellent very good good average or could be improved so please be honest we we do take this feedback we do have a very lengthy debrief after each webinar to figure out how we can serve you the supply chain industry planning better uh also contact information is here on the screen please let us know what future webinar topics you would like to see we have a very long list right now but it's Dynamic we want to serve you and provide you with the information that best serves your immediate needs uh contact us to let us know how we can help you with your supply chain planning challenges well we look forward to seeing you at a future streamlined webinar and again please let us know if there's anything we can do to help you before then very nice results thank you very much I won't share them just to to the uh to be professional about it but uh thank you for the very uh very kind ratings again contact us if there's anything we can help you with and uh I think that's it for today see you at our next webinar soon bye bye