Days Supply with Drive - I need some tweaks!!!!

[davez26]

We've been on Drive a few months, running IRE, and have Days Supply at 28. It seems to work fair, but I think it could be better. ADP recommended we 'run it a bit' before making changes. So here we are.
Without to many details, I have Maintenece, Brakes, Accessories and Other items in seperate sources, (some other sources similiar built for pricing, but same stocking critieria as the 'main part type' source).
Brakes tend to not trend well, and we end up empty from time to time, and we end up OVN.
MNT can get scary low, sometimes, but not always out.
Other items - it's tough to trend that stuff.
Spending some time in the forums, I have been considering a few adjustments this post gave me some thought - wondering what the newest thoughts are with DRIVE?

'Unless ADP has made a very drastic change in the formula for calculating BRP and BSL (looking at a weighted guide where 68% of the guide is in the first four weeks of the formula) you should set the lowest BRP at 28 (four weeks of demand weight) and the BSL to 30. If you use a BRP and BSL under 28 days, the system does not have enough weight in the first few days to create a decent guide and will default to a "sell on buy one" stocking level.' - 'Days Supply' - thread Mike Nicholes, 31May2011.

So:
With ADP Drive, the formula is changed, and defaulted to:
Period Weight
1 .40
2 .30
3 .21
Days in each period = 30
periods = 3
total weight = .91

So the weight is changed.
Right off the bat, I am tempted to modify period 1 to .47 to equal total weight = 1.00 and get more weight to period 1.
But, I also considered something that levels the 3 numbers out, to equal a 3 month trend type of equation. Heck, I could even add more periods to really muck this up!
What seems to work well for Drive?
I am currently seeing Demand filled from Stock ratio pieces = mid 80's
We are smaller, 2 brands on $110K and $40K, so it's really two seperate inventories in one DMS, but they are on seperate Pads and Sources.
Any thoughts appreciated. - Especially from Mike.

[steve@gmdealer]

The default settings on our drive system(which I never have changed) is period 1=.4, period 2=.35 & period 3=.25 so it weighs most heavily on the lastest activities. I know that if you play with demand type, you can up the quanties being ordered. For example, try setting Service Non-Warrany to .6 and the rest to .1 for your maintenance source, run a stockorder and you will see a much different order.

[ScottF]

Do you get daily stockorders?

[davez26]

Daily Stockorders with 4-5 day delivery, ie Monday=Friday, Tuesday=Monday.
I say 4-5 days, as 1/2 the time they deliver after 1:00, after all the action and Demand has occured.

[Richard]

I have to ask, what Brand? Because if you're a GM dealer, most of your time is wasted because of RIM.

[davez26]

Sorry, I am not helping am I?
Jaguar and Land Rover.
Jaguar is my $40K
Land Rover is my $110K.

[partking]

Hey Dave,

This reply is not Drive specific (I’ve never used Drive) but about DMS systems in general.

You mention your days-of-supply setting as 28. I assume you mean your best stocking level or BSL. While that number is important, it is not the biggest driver of fill rate and overall investment level. That bigger driver is really the best reorder point or BRP.

What should be your ideal BRP? That depends on your worst case resupply time. If you submit stock orders daily, you would have a worst time of 3 days (Fri – Mon) till the next order. Now add the time from starting to calculate the order until the filled order is on your shelves and receipted. Maybe 7 days (again counting weekends)? So your worst case is 10 days. So if you reorder while there are 10 days on the shelf, you should never run out, right? Wrong. Demand is rarely linear. Suppose you sell 4 of part X every month. That may not be 1 per week; it could be 2 together every other week. So we need some safety stock. I’ve found a good number to be 2.5 to 3 times the worst case resupply time, more for a smaller store, less for a large one. In the example given, the ideal BRP is between 25 and 30 days.

Now, what’s the best BSL? It depends on who you ask. One day more than the BRP results in sell-one-order-one and minimum investment. Your controller will love that, but it will run your butt ragged. The real number is what you can get away with, compromising between your work load and your bottom line days of supply. Usually a week or two over the BRP is acceptable.

Weighting. You said your system calculates based on 3, 30 day periods or 3 months total with the most recent month counting twice as much as 2 months earlier. That is super volatile both from the weighting and the number of months you look at. The good news is the system reacts very quickly to recent demand. The bad news is that a one time unusual demand will give you a big supply that you may never sell again in that quantity. And forget any seasonal parts. If possible with Drive, I’d look at 6 to 9 months of history sampling, with either no weighting or lightly weighted toward recent periods.

Last, think about another method recommended by Mike Nicholes that he calls “Beyond EOQ.” If you analyze your inventory, you will find that the majority of stocked numbers are stocked in ones. Their BRP and BSL calculate to 0 and 1. You always have a stock out when you sell one, and non linear demand often bites you in the butt. But the good news is that many of those numbers are inexpensive. So you can force a minimum of 2 on the cheap ones and buy a bunch of stock out insurance for not much investment.

Create a custom report that grabs normal stocked numbers with a BRP less than one, no existing minimum and a DN less than $15. Have it add up the cost of one of each number it finds. That total is your maximum increased investment. Re-run the report with a smaller DN value until you get a total you can afford. Then manually add a minimum of 2 (per job in ADP). Now these inexpensive numbers will stock in pairs rather than single pieces. As I said, cheap insurance against stock outs. It’s tedious to set up, but easy to maintain afterwards. Rerun the report monthly to identify newly added parts that fit the criteria.


Hope that helps.

[ScottF]

Im on drive. I have my sources broken down into sales, not parts categories. I have the faster moving items weighted more heavily towards recent history and CP service dept. sales.One thing Ive noticed and there is no emperical formula for its occurance but I see when parts move away from their normal sales trend, they tend to do so beyond the capabilties of the automated system. In other words when a part has a run, it really has a run. Everyday Ill look at stockouts and see if its worthwhile to just add additional quantities based on a relative small and recent window of sales history.

I get daily stockorders six times a week so were in a bit of diffenrent situation than you.