Demand Driven MRP – Strategic Inventory Positioning

Strategic Inventory Positioning

A DDMRP Recap:

There are 5 major elements to Demand Driven MRP (DDMRP):

  1. This page’s focus: Strategic Inventory Positioning
  2. Buffer Profile and Buffer Level determination
  3. Dynamic Buffers
  4. Demand Driven Planning
  5. Highly Visible and Collaborative Execution

“Where” replaces “How much” as the crucial inventory question – as it should ALWAYS have been

Historically, the majority of Planning time and energy associated with Inventory has been devoted to two topics; HOW MANY to hold, and WHEN to reorder.

With Demand Driven MRP (DDMRP ) , there are logical and effective mechanisms to answer the “how much inventory” questions, and to reflect the dynamic nature of the answers; but that’s for later.

Because in DDMRP, the focus is WHERE in the supply chain or within the company to position the inventory.

This question is not trivial;
and the “How Many” question is meaningless
until the “Where” question has been correctly answered.

Choosing the most appropriate locations offers opportunities to solve problems of shortages and unsatisfactory inventory performance and delivery performance, with implications for expediting expenses and for plant productivity as well as for service levels; and, opportunities to improve or gain a competitive advantage in terms of lead time and responsiveness to customer needs.

Correct Inventory positioning can achieve the following:

  • Reduced total inventory
  • Compressed lead time (increased competitive edge, in some environments)
  • Less disruption … a higher tolerance of volatility and variability
  • Reduced “nervousness” in the system, especially if using MRP
  • Increased stability overall
  • Increased agility overall
  • Protection of the performance of critical resources

There are 6 positioning factors that should be considered:

  • Customer Tolerance Time (CTT) – exactly how long is a customer prepared to wait for their product?
  • Market potential lead time – what is the lead time that will enable a company to raise the price of a product, or to win business they otherwise could not?
  • The variable rate of demand – what is the potential for changes in demand – swings, spikes – that could overwhelm the plant’s capacity, stock, or cash?
  • Variability in supply – what is the likelihood of a disruption? What degree of severity is possible? (We know several companies reassessing this question in light of the recent tragedy in Japan)
  • Inventory flexibility, and product structure – where are the locations in the supply chain or inside the BOM that give the most options, and the most potential for lead time compression?
  • Protecting vital operational areas – not just a TOC perspective, where in the system would a buffer reduce disruption to a key resource, or a constraint, or a “Drum” resource (in TOC language)

When it comes to this kind of analysis of anything but a simple environment, computer support should be relied on to perform the heavy lifting.

A concept that needs to be introduced: “ASRLT” or, ASR Lead Time

Anyone familiar with MRP is also familiar with two lead times. Both are extreme.

1. Cumulative Lead Time: This is the longest lead time to buy materials then manufacture the finished product assuming no in-house stocks of anything. (So, unrealistic 99% of the time, for most manufacturers).

2. Manufacturing Lead Time. This is the lead time to manufacture something assuming 100% availability of all components. So, similarly unrealistic 99% of the time for most manufacturers.

What we’re introducing is the ASR Lead Time, or ASRLT, where “ASR” is a carry-over from the earlier shape of the Demand Driven MRP technology – known as “Actively Synchronized Replenishment.”

The ASLRT is variable, and realistic.

For example, if the longest “leg” in the calculation of a cumulative lead time was 30 days, but then inventory was stocked at a certain point (let’s call it “position #1”) along that longest lead time path, the REAL lead time to make the finished product might only be 15 days. That’s the ASRLT. On the other hand, if on that same longest-path, inventory was held at a different place, the ASRLT could be more, or less.

Now, the interesting thing is that of course the ASRLT can be strategically set. For example, if holding inventory at position #1 reduces the ASR lead time to 15 days, then some other “path” might be the basis of that 15 days, and if inventory was stocked along THAT path, the ASRLT might be reduced to 10 days; and, 12 days might be the lead time needed to guarantee winning business from a certain customer or market so the strategic positioning that just gave us a 10-day ASRLT is the basis for winning that business.

Now, using Inventory to compress lead times is fairly intuitive (even if rarely practiced effectively).

But it’s not so obvious that Strategic Inventory Positioning can reduce total inventory in the system. In fact, it can seem counter-intuitive.

The following example is designed to explain the point; but it’s detailed, and nit-picky, and if you’re a big-picture person I suggest you move on to the next topic – Establishing Buffer Profiles and determining Buffer Levels.

A VERY simple example – using Strategic Inventory Positioning to reduce total Inventory Investment

Strategic Inventory Positioning provides multiple benefits – reduced shortages, improved support for schedules, improved service levels, reduced expediting costs, reduced lead times, and reduced system inventories, for example.

For some DDMRP users the inventory reduction may simply be a beneficial side effect, when compared to the value of eliminating the material and component shortages that block production. And we don’t always promote the inventory reduction aggressively as a primary benefit; but the leverage of DDMRP is such that the inventory impact can still be major.

To illustrate how the decision impacts inventory levels, we’ll review a simple example of what in reality can be a VERY complex situation from a manual perspective.

We have to make a few starting assumptions here.

Setting the scene

First of all, we’re going to use the traditional “Re-Order Point” technique as a basis, simply to illustrate the impact on inventory value.

To be clear: DDMRP is a LONG WAY removed from traditional Order Point, it uses a simple but sophisticated replenishment model derived from the Theory of Constraints (TOC) basic model which has been  substantially enhanced by Constraints Management Group, and it could also be considered s fusion of some MRP elements and Lean elements  … but the ROP model is widely understood and avoids the need to explain the Replenishment model first.

Assume we have a part we call “Parent,” which has an average daily usage of 2, a value for Inventory valuation of $1000, and is made from 3 components: Component #1, Component #2, Component #3.

Each component is purchased and has an inventory value of $150. (So, material cost forms 45% of the inventory valuation of the Parent part.)

(Note: in real applications DDMRP can be applied to materials and purchased parts, fabricated or assembled components, and finished goods, throughout all levels of s deep Bill of Material … I’m just keeping it simple here).

The traditional Re-Order Point model is that the Re-Order Point is set to the usage over the replenishment lead time, plus a safety stock. And the Re-Order quantity can be set based on a variety of techniques, one being to set a “maximum” inventory and replenish to aim for that level.

We actually don’t need to know the Re-Order Point for our example; we’re more interested in the “Maximum” stock level. So let’s do this: our ROP model is that the safety stock target is average daily usage over the replenishment lead time; and the Re-Order quantity aims to top the inventory up to a maximum level that is 3 times the safety stock. This is good enough to illustrate the point.


The manufacturing lead time for “Parent” is 10 days.

One component, Component #1, a purchased part, has a 25-day purchased lead time.  This is the longest lead time of the  3 components.

Component #2 has a purchased lead time of 12 days; component #3 has a purchased lead time of 7 days.

If no component parts are “Buffered” with stock …

If no component parts are buffered with stock, then the cumulative lead time for “Parent” is 35 days (10 days manufacturing lead time for part “parent,” plus 25 days purchasing lead time for Component #1.)

If we aim for the safety stock of “Parent” to be average consumption over lead time, we’d be looking here at 70 units (2 units consumed per day, 35 days to replenish). So our inventory “maximum,” which we’ve chosen to be 3 X the safety stock just to keep things simple, would be (3 X 70 units) = 210 units.  With an inventory value of $210,000 ($210 units at $1000 each).

If we Buffer for Component #1 …

If we decide to hold a stock buffer for component #1, the inventory for component #1 will of course increase from the starting inventory of zero. If we apply the same Re-Order Point model, the safety stock would be 25 days X 2 units consumed per day … 50 units. So with the maximum inventory set at 3 X the safety stock, we’d be carrying 150 units. And with the individual Components assigned an inventory value of $150 each, the value of 150 units of Component #1 will be $22,500.

BUT: the impact of the decision to hold a Buffer Stock of Component #1 is that we can hold less Inventory of “Parent,” while still maintaining or even improving the availability of the part “parent.” The rationale is probably obvious to you: if we’re buffering for Component #1 then the cumulative lead time of “Patent” is now 22 days, (10 days manufacturing lead time plus the 12 days purchased lead time for Component #2), not 35 days.

And that means that the safety stock for “Parent” would be 44 units (2 consumed per day, 22 days to replenish) instead of 70 when there was no buffering; the target maximum inventory for “parent” is therefore 3 X 44 units, because we chose 3 times safety stock as our target maximum; or 132 units.

This is 78 fewer units of “parent” than when we did not buffer Component #1.

So we have been able to reduce the investment in Finished Goods by $78,000.

For a Net reduction in inventory value of $55,500 (reduction in “Parent” inventory of $78,500, netted against an increased $22,500 inventory of Component #1.).

With NO REDUCTION IN THE AVAILABILITY OF THE “PARENT” PART. This is important. By making a strategic decision to hold inventory of Component #1 we improved the response time to the market from 35 days to 22, we reduced the inventory investment by $55,000, and we maintained the availability of “Parent” to the market. (We may even have increased it … see later).

That’s not a bad combination.

Interesting What-Ifs

Now … what if Component#1 wasn’t just used-on “Parent?” What if it was used on 10 other “Parent” type parts? What would the potential Net reduction in Inventory be?

What if we were looking at 200,000 component and finished goods records, instead of 4? What if we were  looking at finished products built in a 15-layer deep Bill of Materials, instead of 2? What if the value of the finished unit was $100,000 and not $1000?

A little more sophisticated …

In fact,  from a statistical standpoint, the more used-on situations for Component #1, the lower the inventory of Component #1 would need to be to maintain the same degree of availability for the Parent parts. In other words, it wouldn’t need $225,000 of inventory of Component #1 ($22,500 per “parent” times 10 “Parent”-type parts) to provide support for 10 Parent-type parts with identical characteristics to Parent.

And, with reduced cumulative lead times for “Parent,” there would almost certainly be a reduction in the variability of the supply of Component #1, probably justifying a reduction in the safety stocks of the “Parent” products.

So the Net reduction of inventory would improve as a result of both these mechanisms … while availability increases and inventory levels drop.

And this is before we add the impact of the Dynamic Replenishment model of DDMRP.

The Demand Driven MRP  model enables the DDMRP user to set and maintain the Buffer Stock levels with a great deal more sophistication than described above, making the impact on Parent part availability and inventory levels even more advantageous.


Be Sociable, Share!