Setting Customer Service Levels… Without Managing Stock Levels

Your customers want short lead-times and reliable on-time deliveries.

Alas, this is not always possible.  Serve every customer to the standard they ask and you might quickly go out of business.  Inventory and capacity are constrained.

You make choices about setting the right service level to each customer.  Without a good policy, many manufacturers choose to hard allocate inventory upon order.  This can result to inflexible supply and a lower overall service.

The smart manufacturer is able to set policy to ensure that the most profitable customers receive the highest level of service.

When supply chain managers think about service level, they go straight to inventory sizing and days of supply.

When operations managers think about service level, they should go straight to reducing lead-times and improving response.

What if I told you that there was a way to improve service levels before you do any inventory sizing, capacity optimization or waste reduction?  And at the same time, it can help solve the inventory allocation problem that many companies suffer?

Interested?  Read on.

Some Customers Are More Equal Than Others

The Pareto Principle shows it is likely that customer profitability follows a power law.  The most profitable will be many times more profitable than the least.

We are suspicious that some customers actually lose us money.  We  know that others can make or break the company.  Service discrimination is a fact of business life.

So, it is a good idea to have a clear view on the value behind each customer relationship.  Create tiers or classes of service.  Execute orders according to a defined service level.

Inventory Allocation and Planning

Most discussions on service level begin and end with inventory sizing.   This is a shame; ideally, stock levels should be set against production lead-time, capability and variability of demand.

In practice, it is not easy to set production capability against customer service.  So, inventory is usually involved.

There is a whole conversation (series of articles?  workshop?) to be had on how to size inventory according to demand requirements and the working capital plan.

This is not that article.

This something that arguably comes before sizing :  It is allocation.

In a previous article, I wrote about the impact of allocating inventory upon receipt of order.

Many Make-to-Stock companies will check on-hand and on-order materials upon receipt of order.  Some go further and “hard allocate” the stock so that it cannot be used for other orders that may be received later.

This is contrast to a flexible FIFO method, where material is allocated to an order according to the ship date (or due date in manufacturing).

Hard allocation of stock is not a problem if every order has the same lead-time.  This means the sequence of orders received is the same as the sequence of orders filled.  The result is the same and as flexible as using the ship date.

With variable lead-times, some problems arise.  If I assign inventory upon order, I may be ignoring incoming supply which arrives early enough to satisfy this first order, but too late to fill an order received afterwards but is due sooner.  This article has an example, plus a download Excel tool to make the comparison.

If you download the tool, you will be asked to complete a survey.  We asked 2,160 readers (and counting…) the following question:

The results were somewhat surprising to me.  I have been solving  problems caused by hard allocation for over ten years now. (and I am by no means long in the tooth, yet).  It is not exactly a best practice for supply chain management.

Yet, behold the data:
It shows that almost half of the respondents hard allocate everything, and over 75% regularly practice hard allocation.

The download tool (you also use the survey link above) shows that a flexible method based on FIFO wins almost every time over hard allocation.

So, I hope that some of you are thinking of ways to loosen up inventory and make it more flexible.

This doesn’t mean that 75% of readers to this blog use hard allocation of inventory.  The sample is still quite small (2,160 total downloads from over 21,000 visits a month).   The likelihood of someone wanting to download a tool to analyze the effects of hard allocation is higher if that person uses that method.

Nevertheless.  It means that many manufacturers are doing it.

Why?  The most common reason is a desire to look after certain customers.  After all, hard allocation is a pretty safe and simple way to ensure that the most important customers don’t get short.

Demand Allocation and Service Level

Allocate stock to some customers upon order entry (and from next available supply if no on-hand stock).  Allocate stock to other customers upon order due date (and manage shortages by projecting total supply).

This sounds simple.  It is not quite as simple as it sounds to do the planning in Excel.  However, in the interests of improvement and better customer service for all, I present to you my spreadsheet model for service discrimination by inventory allocation.

Before you download it, you might find interesting the method with which it was built.

This tool stacks together sales orders, supply orders and inventory.  And then makes 3 comparison projections on the service level based on 3 different scenarios.

The first projection is based on hard allocation at order date.  The second is for a FIFO scenario at order due date.  And the third is for service level scenario that does hard allocation for some customers and FIFO for others.

In this third scenario, customers for hard allocation are more important.  We model this using a points system.  For example, if we get 6 points for an on-time delivery for important customer and 4 points for others, we can see how well we have done overall.  This takes the customer importance into consideration.

All of the stock calculations are done in a single worksheet.  It performs in seconds and uses a sort method that can be used to power very fast analytical calculations.  Sort is a very fast and flexible way of calculating large tables of data.  This comparison tool has some good inventory projection techniques packed into one worksheet and a fast routine.

This tool generates sales order quantities randomly.  It means you can run it multiple times to see the result with different demand.

The results are interesting.  It seems that if your upper tier of customer is twice as important (and profitable) than your lower tier, then you will get approximately 10% improvement over all FIFO allocation and 14% improvement over hard allocation.

Just like before, FIFO always beats hard-only allocation.

So, play around with this comparison tool with different numbers.

Leave a comment below and let me know how this simple model compares with your inventory allocation method.   Any other reason for using hard inventory allocation, aside from service discrimination?