John Wilder Tukey (June 16, 1915 – July 26, 2000) was an American mathematician with a Ph.D. in mathematics from Princeton University. He became founding chairman of the Princeton University statistics department in 1965, even though he was dividing his time between his obligations and Princeton and his invaluable work at AT&T Bell Laboratories.
As a scientist and mathematician, you might think he would be all about being “precise.” Nevertheless, he strongly affirmed this:
"Far better an approximate answer to the right question, which is often vague, than an exact answer to the wrong question, which can always be made precise." 
Getting Exact Answers to the Wrong Question
Traditional ERP systems—regardless of the price paid for them—almost always embed traditional MRP (material requirements planning) for use in manufacturing and supply chain planning and execution.
- WHAT you should make, buy or transfer
- WHEN you should make it, buy it or transfer it
- HOW MUCH you should make, buy or transfer
It does so with the very best of intentions—by the way.
The goal of MRP systems is to keep you supplied with WHAT you need, WHERE you need it, and WHEN you need it.
There are a few problems, however.
One big problem is that MRP doesn’t care if you need to attempt to bend the space-time continuum to meet its calculated plan.
Another huge problem is traditional MRP is dependent upon, typically, tens of thousands of data points, settings, and configurations in the ERP system—and some of those settings were “best guesses” when they were configured. Others were pretty accurate when they were set—six months or six years ago!
The results coming from traditional ERP/MRP planning systems are typically not something upon which the supply chain managers and inventory replenishment planners can execute.
This is precisely why most companies—way over 90 percent of companies—that have access to, and (may) use, traditional MRP, supplement the data coming from the ERP/MRP systems using spreadsheets, homegrown databases, and even whiteboards or clipboards before making final decisions.
The data coming from the traditional ERP/MRP system are very precise—and almost always precisely wrong.
Approximately Right Answers to the Right Question
Demand-driven MRP (DDMRP) is a learning system (meaning it is constantly updating itself with the latest relevant data coming from your ERP system). Its simple, high-effective signals give your supply chain managers, production planners, inventory folks and any others the right answers to the right question.
The RIGHT QUESTION is this one:
What should our priorities be in order to assure the FLOW of relevant materials in our supply chain right now?
Of course, those signals must lead to the relevant actions, and that is done simply enough. A simple set of signals (such as those in the accompanying figure) tell everyone the required priorities and what they need to know to take appropriate actions.
Whether these SKUs are make items or buy items, it is easy to see that SKU #0060480 is in the RED ZONE with only 2.25% remaining in its (stock) buffer. This buffer requires immediate replenishment and the quantity (to make or buy) 1,247 units.
If all of the RED ZONE matters are attended to, then the YELLOW ZONE SKUs should next be addressed. SKU #0068565 requires replenishment of 4,797 units.
What’s the Difference?
DDMRP doesn’t try to incorporate tens of thousands of variables—with a high probability of error—and then attempt to tell you precisely what you should buy, make or transfer, at precisely what time.
Instead, DDMRP give you simple signals that help you understand…
How well are my various buffers protecting FLOW (read: profits) right now?
If a buffer is in the RED ZONE, it needs immediate attention if it is going to continue to protect the FLOW of relevant materials in your supply chain. The lower the remaining percent of buffer, the more urgent the attention is required. Thus, priorities are easily determined.
Don’t Confuse “Precision” with “Accuracy”
Traditional MRP’s numbers are always precise, but they are seldom accurate. DDMRP’s numbers are far more accurate in terms of supporting the FLOW of relevant materials.
Robert E. Kass, of Carnegie Mellon University, says, “[T]here is a simple and elegant way to combine current information with prior experience in order to state how much is known…. It makes full use of available information, and it produces decisions having the least possible error rate."  [Emphasis added.]
Isn’t it time that you and your supply chain found a truly simple and very effective way to help assure the FLOW of relevant materials across your supply chain? (By the way, mass shipments and mass production of goods are not needed at the present time is the FLOW of irrelevant materials and actually hurts the bottom-line of everyone in your supply chain.)
We can help.
Please leave your comments below, or feel free to contact us directly, if you prefer.
 McGrayne, Sharon Bertsch. The Theory That Would Not Die: How Bayes' Rule Cracked the Enigma Code, Hunted Down Russian Submarines, and Emerged Triumphant from Two Centuries of Controversy. New Haven: Yale University Press, 2012.
 By the way, there is book entitled Approximately Right, Not Precisely Wrong – Cost Accounting, Pricing & Decision Making on a different topic than I cover here. Nevertheless, you might find it interesting.