By using this website, you agree to our Terms of Use (click here)
Hello AUG Community!
Our org is currently undergoing a transition in one of our businesses from primarily make-to-order using Kit Assemblies into a majority of make-to-stock (with some MTO) using Acumatica's Manufacturing Module.
We have great help from our partners in all the implementation requirements, but I'm hunting for best practices on how our Merch team should interface best with the system to incorporate demand forecasting for our Direct to Consumer business. The MRP seems like it's great for B2B companies who take large pre-orders and can plan out their business from sales order demand, but I'm a little bit at a loss for what the most efficient way to use this system is when everything is based from Demand Forecasting, not existing SO Demand. Yes, there are mins and maxes, and safety levels, etc, but in our business we map out our forecast at the week level, meaning to have the best inventory management possible, these numbers would/should change weekly, or at the least by period/month. We have our own fairly robust forecasting methods outside of Acumatica we've relied on for years, but we need to make significant modifications to them to translate everything into a MTS type business. It's possible to do, of course, but I'm trying to investigate best practices and see how other people are using the system before we forge ahead on our own.
Is there anyone else out there in a D2C e-commerce space who relies on manufacturing? Would love to chat about your forecasting processes and how you best use the system!
Thanks all!
I don't have any direct experience with this scenario so I can only provide some theoretical thoughts and questions:
1. Your company produces seeds so you probably don't have shelf-life concerns limitations so you main limitations would be cash and storage space right?
2. What's driving the need for make to stock? Is it outages? Is it a desire to include a greater number of SKUs in your forecasting? Is it something else?
3. To make the Acumatica Manufacturing module happy, could you simply enter a dummy Sales Order with a customer called "Make to Stock"?
We're in the process of deploying MRP/Manufacturing. We're setting up do deal with print on demand and the management of annual stock plans. I've already worked through the process for forecasting our stock plan and have opted for the following, which might be a consideration for you...
We opt to enter in a manual Forecast based on historical sales and consideration of e past, current, and future demands of these products, there's a little tea leaf reading in there too... all of this is done externally from Acumatica BUT does leverage some GI's to keep the evaluation relevant.
Please appreciate that I'm going by memory, a little rusty since we had a slight change in focus for the past few months, but we're about do dive deep into this again.
Considerations
When we work on our stock products we like to focus on one product at a time for a duration no less than one week, usually 3 or more.
A single production may be produced a few times over the course of the year as we alternate between products in a timely fashion and with the hopes of aligning with the possible demands.
Our steps...
Open our main Excel do, which leverages a few GI's
Evaluations and preparation are performed within Excel.
Read some tea leaves and establish which product will be worked on when, in one week increments
Enter a record for each item that represents a wave of activity indicating the date of production and the total quantity.
Results...
With Acumatica being set to chop up the forecast demands into weeks it all seems to plan things out quite well. Each record entered into the Forecast screen will be split up into Production orders, scheduled one per week.
Trick...
Oh! the other trick... have a separate warehouse to perform all of this in so when MRP evaluates everything it simplifies things since it will only evaluate the quantities within that warehouse... ie "STOCKPLAN".
I'd be happy to share where we're currently at to give you other insights. let me know and good luck! 😉
Thanks for the responses, Tim and James!
Tim, to response to your comments:
1. Seed does have a shelf life, because as we store seed germination rates go down - albeit relatively slowly, but we don't like to hold on to seed for more than a year. If we do keep it longer we send it off for testing to ensure germination is strong, and so we can keep seed longer as it'll usually keep a bit later. But more than that yes, it's good just for cash management to keep things tight.
2. The drive to make to stock is for capacity. We have a 10 week period where we might ship out something like 50% of our entire annual business. We ship year round, but in spring and in fall we have a short window where everyone wants everything all at once. It's like a gardener's christmas season! So in light of that, in order to maintain a 48 hour turn time to customers, we have to make to stock all of our seed bags.
3. We could do dummy sales orders, but that would really muddy up the system. The MRP has a forecasting ability to input demand, which I'd prefer to use rather than dummy SOs, but the problem there is that the forecast changes every week as we get closer to the season.
James, I appreciate your writeup there! I think one main difference in our systems is that for us, instead of working on one item per week, we might be working on 10-15 items or more per day.
Enter a record for each item that represents a wave of activity indicating the date of production and the total quantity.
What do you mean when you say enter a record? Is this in the MRP Forecast utility?
I think we have a pretty good system outside of Acumatica as well - our 'tea leaf' system - haha, but doing my best to see where we can streamline and keep manufacturing folks in acumatica as much as possible without going to external docs.
As I investigate more, I'm starting to think we might start off by doing the initial planning outside of acumatica, but then having our merch team use Velixo or something similar to directly import Minimum order Quantities weekly or periodically that the manufacturing team will then just rely on in system.
In an ideal world I'd be able to upload a 'plan' that could serve to build out, and with help of advanced planning module, spread out, work orders based on our system inputs. Those orders could be tied together with some attribute that they are 'plan' derived.
Then as we get closer to execution of those orders, the system could ingest demand curves and offer 'forecast' work orders in addition to or in place of the plan orders. I know that's asking way too much of the system, and as with anything else, it would be garbage in garbage out, so you'd also need a robust method of ensuring all of your data is complete and accurate constantly. But that's basically what we do outside of Acumatica. We have a plan, set work to that plan, then as we get closer, 4-6 weeks, we start revising it to a forecast which as we get closer and closer is really just pure demand curve based.
Oh, and thankfully for us, all of our manufacturing is in the same warehouse, James, so we don't have to worry about that limitation!
Re : What do you mean when you say enter a record? Is this in the MRP Forecast utility?
Screen ID : AM202000 allows you to manually add in a record that represents a demand being made on the system to produce an item. In this example, it will create a Production Order for 4000 units for each week found between the Begin Date and End Date. You can also import these records rather than manually entering them.
Note that managing one item per week or 100 distinct items doesn't really matter since you're doing all the planning outside of Acumatica. I've ran through a model where we had the entire year's stock plan all ingested at once and it did a decent job rending all the production orders and, in some cases, we doubled up on the products being produced and it all worked out well. Consider that, at this level, it's not actually evaluating much and is simply following an algorithm based on the information presented in the Forecast screen.
Re : I think we have a pretty good system outside of Acumatica as well
absolutely get it and we're likely to do the same and may never be able to, practically speaking, leverage the built in forecasting feature BUT it could be something to aspire to.
Re : In an ideal world I'd be able to upload a 'plan' that could serve to build out...
You could do this now with a combination of a nice Excel doc and linked GI's. That's how we're approaching our planning of stock. Are you any good with GI's and Power Query? The Excel doc would also contain various thresholds based on managerial preference. This is what we're doing. Heck, you could kick it up a few more notches and even push it through Power Bi and do some wild stuff that could continue to monitor all and potentially even build out predictive models and leverage AI, That's where I'm going next lol.
Re : Oh, and thankfully for us, all of our manufacturing is in the same warehouse...
A warehouse is nothing more than an organizational unit within a database. the system doesn't actually know what a warehouse is. it only knows what tables, fields, relationships, and such are and such. When MRP evaluates the demands it evaluates products in the same warehouse and adjusts based on its findings.
Scenario One
- Make everything in the same warehouse you're selling from... MAIN
- There are currently 500 units in MAIN
- You've entered a demand, a record, in the Forecast for 4000 units.
- MRP will consider the total quantity currently in that warehouse and subtract it from the demand since they're already there resulting in a production order for 3500 units. This might be what you want or maybe not...
Scenario Two
- Production exists in a warehouse other than where we sell from., let's call it STOCKPLAN
- MAIN contains 500 units
- You've entered a demand, a record, in the Forecast for 4000 units.
- MRP will consider the total quantity currently in that warehouse and subtract it from the demand but since STOCKPLAN is only utilized for planning of the stock then there's never anything in it, outside of WIP. This will result in a production order for 4000 units. This might be what you want or maybe not...
FYI, I have a saying... "I'm no expert but I'm certainly happy to share my experiences and all I ever hope for is that someone will take what I share, mix it into their own pot of soup, and see what comes of it, it just might inspire you to the solution you're lookin' for..."
😉