Volume discounts have been an assumed part of business for s long few people stop to think of them as the epitome of un-lean thinking they really are. It should be pretty obvious – telling customers we will charge less if you buy more than you need and keep it in inventory.
The boys writing for the Harvard Business Review blog are rarely relevant and even less often right … and when they are it is usually for all the wrong reasons. “When It’s Wise to Offer Volume Discounts” is no exception. While the author is right when he says, “While there are sound reasons to provide volume discounts, most companies over use them and offer them with too little thought,” he misses the most fundamental point. That is, as set ups times go down, the logic for volume discounts collapses.
It is really a matter of common sense that production works best – ‘best’ meaning ‘cheapest’ – when level loaded. Nothing good comes from demand that flows through a factory like a pig flowing through a python. Yet that is precisely what he suggests as his fourth instance in which volume discounts are a good idea:
“To encourage a large order instead of a series of small ones: Pharmacies often offer a discount if you purchase a year-long supply of a common prescription, for instance, instead of filling it monthly. This price break yields higher profits as pharmacies only have to incur the costs of filling a prescription once, instead of twelve times.”
The costs of filling a pharmacy are basically the same as the costs of making anything, which means they are complicated. Most of them are fixed and the costs to make something, including a prescription, are minimal – the only real variable cost is the cost of the materials – the pills themselves. The justification for volume discounts is based on the tradeoffs of competing ways to waste capacity. The author – and traditional accountants – seem to think that wasting production capacity by spending it making more than customers need is better than wasting it on setting up the operation.
If it takes the pharmacist 1-1/2 minutes to set up the order and 5 seconds per pill to fill it, an order for 30 pills takes 4 minutes. He can fill 15 prescriptions per hour. The old logic is based on the concern that such an operation would waste 37.5% of the pharmacists time on setting up prescriptions = 15 orders per hour X 1.5 minutes each = 22.5 minutes per hour ÷ 60 minutes = 37.5%.
Far better the old school thinker will tell you to fill orders of a year’s worth: same 1-1/2 minutes per prescription, but 360 pills. You can do the math and see that this has the pharmacist spending 97.5% of his time on the pills and only 2-1/2% setting up, compared to the 62.5% when he fills prescriptions a month at a time.
The math is great … so long as the only thing you care about is direct labor – the cost per pill X the wages of the pharmacist; and, of course, if you make the assumptions that pharmacists are a variable cost and you assume inventory is an asset with no carrying cost; and if you don’t care about customers too much.
Those things can’t be assumed, however, and you need to think about the following:
Instead of keeping 30 pills in inventory for the maximum number of folks who might wander in for a prescription of crapozene you have to keep 360 for each of them – 12 times a much inventory. And, of course, this is X a very big number because the pharmacy has lots and lots of different pills it offers. More inventory means more shelf space, more shelves, more bins, more inventory errors, more time to manage, maintain and keep track of all that inventory.
There are exaggerated labor problems, with ten customers coming on for prescriptions on Monday, but only six on Tuesday. That means the pharmacists has 10.25 hours of work to do on Monday – overtime – but only 6.25 hours of work to do on Tuesday – idle time.
And, of course, customers have to wait longer depending on where they are in the queue. A little over an hour per customer for each customer ahead of them when the prescriptions are filled for a year’s worth, instead of 4 minutes if the prescriptions are filled for a month at a time.
But it is hard to put numbers on capacity imbalance, inventory costs and poor customer service, and it is easy to calculate labor cost per pill, so labor cost dominates the decision.
What if the pharmacy is a lean one, though? What if the pharmacist read Shingo’s SMED book and has his set-up time is negligible? Then all of this math collapses and the cost per pill is just the 15 seconds each – no matter what the quantity.. Encouraging bigger orders just creates all of the inventory, capacity waste and lousy customer service with no off-setting benefit at all.
In the end, volume discounts only make sense for lousy manufacturers – and then only because they don’t really understand their costs. For a lean manufacturer, volume discounts make no sense at all. They are just begging customers to create production problems for them.