In thise case dummy-customer-specific forecasts would make sense right? We use Smart/IP&O to get granularity at customer/site/part level…but summarize it at the site/part level within Kinetic. We’ve got too many customers ordering 1s and 2s a month and we don’t need that cluttering things up.
We’re using Smart/IP&O to do the forecasts at the granular customer/site/part level but we’re rolling up to site/part within Kinetic.
ok - been reading this for a while and I want to offer something different, maybe simpler to manage. I’ll say I haven’t thought it out completely, but that’s what this thread is for, right?
Since you have a dummy order, for a dummy customer, could you not flip this around and create a dummy Part with a min On Hand qty of say 5 and a build qty of 1? You could really bypass most of the rest of the overhead work, and just sell it on a REAL order when you determine it’ll be sold (either during or after manufacturing). You are putting it into inventory anyway, so the $$ is going to the right place.
And MRP will still get all the demand lined up for you.
No I think customerleas forecasts cause they don’t know who. They just want the demand. Though maybe they will need to in order for other orders (MTO) to consume it
Since we went off the rails a while ago, someone want to give a quick high level of how mrp works with stock jobs? How would we even start with demand? Am I even asking the right questions?
You just enter a Job to Stock (no demand) then MRP will create suggestions for other pieces etc
If Demand Management’s an option…demand contract/entry? There’s an option for forecast/unfirm/firm.
Went to a 20-minute meeting and there are 14 new replies in this thread, plus the other 2 dozen I didn’t reply to.
Working on it…
And now we’ll all run off to lunch, and pick it up again in an hour or so 
You don’t need to do Customer specific Forecasts, you can just do a Forecast for a part and the system will still consume it when orders come in.
@Mark_Wonsil said no dummies.
Possibility, but still seems like extra work instead of a generic part forecast.
OK, Forecasts.
Again, the pro is that real orders are able to “consume” the forecast demand. Cool.
Well, what if you don’t sell the thing that you forecasted?
We are engineer-to-order. We really don’t sell the same truck twice. Forecasting only works if you sell the part number in the forecast.
I mean we will sell 2 or 100 of the same truck on a single order to a customer. But if they want that same truck next year, they cannot get it. We’ve added some standard safety feature since last year or something. And we give it a new part number.
But even then, we don’t really sell “standard” trucks. There is always some customization from the customer. Headlights, grating color, mud flaps, their logo decal, etc.
Right, that.
Again, we don’t use a configurator (but we do have the license). I don’t know if that would help or not.
Probably not as they are moving to pushing the BOMs from the CAD software here.
I’m very confused by this, you are saying you want to make some trucks to stock but also you don’t do this?
If you don’t sell the stuff you forecasted it sits in inventory until you sell it
I don’t know if this is critical to management, but you can’t exactly schedule that, right?
Even with these extra trucks, we still need to plan ahead for them.
But IDK, maybe I’m thinking about this too much in the way we do it now. Honestly I don’t know what exactly drives the decision to make X stock trucks or X+3 or whatever number.
What is the final destination of these stock trucks. Are they just sold as is? or are they too special customized in the end?
If so then it sounds like you are building a partial assembly ahead of time (a basic truck) and then later on when a customer actually orders a new truck, you can use that “basic truck” but add wings, or windows, or extra wheels… (I clearly know nothing about trucks)
Is that the case?
I was going to ask the same.
Is it a finished truck that goes into inventory or an almost finished truck that is missing some parts?
Similar thought here too. Is there potential for a divide and conquer solution? If 80% of the same truck is repeatedly sold, that’s forecastable. Similar for options, if they’re reasonably consistent. Would it be possible to base forecasting on the subset of what’s consistent and incrementally sneak up on portions of the rest with a different or improved solution?
That is the new Planning BOM functionality they recently released. But yes.
What I mean is,
So 754000 does not consume the forecast for 753999.
Good questions.
This is where it is really hard to step back from working here 12 years and be objective.
