TOC in Distribution · Lesson 5 of 8

Sizing the buffer

Four factors decide how much stock a location needs. Only one of them is really in your control, and it changes the other three.

12 min · Not completed yet

The four factors

How much stock a location needs depends on four things:

  1. Level of demand: how much sells on average.
  2. Replenishment time: how long it takes, from a sale, until that stock is replaced on the shelf.
  3. Variability in demand: how much sales swing around the average.
  4. Variability in replenishment time: how often supplies arrive late.

All other things equal, stock needed rises with each of them.

The rule for the buffer

A buffer has one job: to keep the product available until the next replenishment arrives. So it must cover the maximum demand that can reasonably happen within one replenishment time, not the average.

Picture daily sales laid out in buckets, each bucket one replenishment time long. Some buckets are quiet, some busy. The buffer must survive the busiest bucket. The word maximum is in quotes on purpose: not the wildest day in ten years, but the busiest stretch you can reasonably expect. As good managers we should be paranoid, but not hysterical.

Buffer = “maximum” expected consumption within the replenishment time, adjusted for how reliable supply is.

Try the four factors yourself:

Notice the warning that comes with this rule: if you size buffers on the maximum demand within today’s long replenishment times, you will end up holding a very large amount of stock. The rule is right; the replenishment time is the problem.

Which factor can you actually change?

Pause and think: you want to need less stock. Which of the four factors is really in your control?

  • Level of demand: sell less
  • Variability in demand: ask customers to buy more evenly
  • Variability in supply: make every supplier perfectly reliable
  • Replenishment time

Selling less is not an option, demand swings are the market’s nature, and making every supplier perfectly reliable takes enormous effort. That leaves replenishment time, which is largely made up of your own ordering and shipping habits.

The hotel linen

Think of a hotel and its bed linen. Every day, guests use fresh sheets and towels; the used ones go to the laundry and come back after a while. How many sets does each room need?

If the laundry takes a week to return the linen, the hotel needs at least seven sets per room, plus a few for days when the laundry is late or the hotel is full. If the laundry returns it the next day, two or three sets are enough. Same rooms, same guests, same demand. Only the replenishment time changed, and the inventory fell drastically.

Replenishment time is the key

Replenishment time has a direct effect on the buffer: it multiplies everything. But it also has an indirect effect, which is even bigger:

  • The longer the replenishment time, the bigger the swings in demand you must cover. Over three days, demand varies a little; over three months, it can double or halve. A long horizon makes every forecast less accurate.
  • The longer the replenishment time, the more variable it gets. Long, complex supply routes have more ways to slip.

And it has one more ramification, worse than all the stock it demands: a long replenishment time takes away the system’s ability to react. Short seasonal peaks, a sudden jump or slump in demand, a market shock: if replenishment takes two months, by the time new stock arrives the moment has passed. You are stuck with the wrong stock for the new reality.

Check your understanding

Answer all five. Get four right to complete the lesson.

1. A buffer must cover…

  • Average demand in a month
  • The maximum reasonable demand within one replenishment time
  • One day of sales
  • Last year’s sales

It has to survive the busiest stretch until the next replenishment arrives.

2. Which of the four factors is most in your control?

  • Replenishment time
  • Level of demand
  • Variability in demand
  • Weather

Much of it is your own ordering and shipping behaviour.

3. Besides needing more stock, a long replenishment time…

  • Improves forecast accuracy
  • Reduces variability
  • Increases the demand swings you must cover and makes supply less reliable
  • Has no other effect

Longer horizons mean bigger swings and more ways for supply to slip.

4. Why is a long replenishment time dangerous when demand suddenly changes?

  • It isn’t
  • Because prices rise
  • Because distributors complain
  • Because the system can’t react before the moment has passed

By the time new stock arrives, demand has moved on.

5. Halving the replenishment time, with everything else the same, roughly…

  • Doubles the buffer needed
  • Halves the buffer needed, and also reduces the swings to cover
  • Has no effect on the buffer
  • Makes forecasting harder

Direct effect and indirect effect both work in your favour.