Imagine, if you will, a supply chain where the repeated sequence of events is like this:
- Consumers buy products (day 1)
- At the end of the day, the retailer send the shelf take-away data to the distributor
- The distributor ships the quantity consumed (yesterday) to the retailer (day 2)
- At the end of the day, the distributor sends shelf take-away data to the wholesaler
- The wholesaler ships the quantity consumed (the day before) to the distributor (day 3)
- At the end of the day, the wholesaler sends shelf take-away data to the manufacturer
- The manufacturer ships the quantity consumed (the day before) to the wholesaler (day 4)
The retailer would need to keep an inventory of an average day’s demand, plus some quantity to cover variability in daily demand (assuming zero variability in supply—just for this example’s sake). Each participant in the supply chain would also need to keep on-hand a quantity equal to an average day’s demand (at their aggregate levels), plus some quantity to cover variability in demand.
Forecasting would be virtually eliminated. I say, “virtually,” because if artificial demand variability is introduced into the supply chain through policies and procedures—like, short-term price promotions or quarter-end sales incentives for the salespeople or channel participants, then, of course, the impact of those short-term changes in demand must be estimated (forecast) and accounted for in the ordering and replenishment process.
Lots of agreement
There is lots of agreement that this approach to supply chain management would work—in theory. There is also lots of agreement among supply chain managers as to why this approach is “impossible to achieve” in most circumstances.
Let’s agree that this idealistic supply chain may, in fact, not be achievable in most real-life circumstance.
Back to our current reality
But, let’s look at how most supply chains today actually function (or, fail to function, in too many cases).
This supply chain look quite similar to the idealistic supply chain above. That is because what are not shown in this simple diagram are all of the trouble-causing delays introduced into supply chain execution.
Here is what the steps look like:
- Consumers buy products over days 1 to r
- Day r +1, the retailer places an order with the distributor
- The distributor accumulates actual demand over one or more days (d)
- On day r + d + 1, the distributor places an accumulated order with the wholesaler
- The wholesaler accumulated actual demand over one or more days (w)
- On day r + d + w + 1, the distributor places an accumulated order with the manufacturer
- The manufacturer may also accumulate orders over 1 or more days (m) before producing the products
- On day r + d + w + m + 1, the manufacturer produces the required goods
Since consumer shelf take-away data probably does not flow end-to-end across this supply chain, the manufacturer that produces the goods is separated from knowledge of actual consumer demand for its products by, at least, r + d + w + m days, plus any additional delays induced into the order process by excess inventories being held anywhere in the supply chain.
If these numbers are typical, they might look something like this:
- At retail (r) = 1
- At distributor (d) = 7
- At wholesale (w) = 14
- At manufacture (m) = 14
- TOTAL INDUCED DELAY = 36 days
This, in itself, is problematic.
How do most address this situation in the supply chain?
Unfortunately, for a great many small to mid-sized companies—companies with sales, perhaps, into the hundreds of millions of dollar—the answer to this problem is sought primarily in finding ways to improve forecasts and optimize inventory quantities to cover over the days of delay and the negative effects introduced into the supply chain.
Companies appear to be willing to spend hundreds of thousands of dollars to carry more inventory and to purchase more and more advanced analytics in order to produces more and more sophisticated “guesses” about what their market will actually be like 30+, 60+, 90+ or even 120+ days into the future.
Is this really necessary?
We do not believe that this is really necessary in a great many cases.
Maybe you can’t cut your supply chain replenishment cycle from where it is today to a daily cycle. In fact, it is likely you cannot—and should not. However, it might be a very good step to produce a plan for improved supply chain collaboration to cut your current supply chain end-to-end cycle in half over the coming six months. Then, perhaps you can look at halving that cycle-time again six months or a year later.
Much of what is necessary to make such changes are low-cost, or even no-cost, solutions.
We think these options deserve more attention.
Please let us know your thoughts on this matter by leaving your comments below.