Batching – An Enemy Disguised as an Ally

For-profit businesses have a simple task – make profit. Well, maybe not that simple to attain, but simply put for a business to make profit all it needs to do is to generate revenues that are greater than its expenses (and if the business is interested in sustaining the profit making, it needs to ensure that it grows, and therefore that its revenues grow faster than its expenses).

The nature of these two forces is quite interesting, and one of the characteristics they have is absolutely opposite. While revenues generation is subject to high uncertainty and somewhat out of the business control, cost is highly certain and almost fully under the business control. This core difference leads to a fundamental difference in the way management treats the two; we hope for revenues and we ache costs.

Funny thing, we tend to be much more attuned to our aches than to our hopes, and sadly we tend to focus our managerial processes and tools on “optimizing” cost. You will notice that I placed optimizing in quotes, as I do not believe the term is relevant at all for the business world. And just for putting this argument aside, if you accept that businesses must grow it will immediately mean that there cannot be an optimum value for any of the relevant measures. But, let’s go back to our tendency to focus on cost and try to better understand the effect of that on our business overall performance. As this is a very large subject, I chose for this article to discuss BATCHES and BATCHING.

Batching is all around us, we use it everywhere in our business lives as well as in our private lives. We were educated and we firmly believe that batches give us the most cost effective approach. Please note the term – cost effective. Cost is the leader, effectiveness follows. How do we measure this cost effectiveness? Do we measure it by the effect the batching choice has on our overall cost? Well, mostly we do not; we rather measure a different measure and believe it gives us a good indicator to the effect on the overall cost. I am sure you are familiar with this measure; it is called “cost per unit”. So we batch as the cost per unit is lower vs. if we did not batch. And we believe (hope?) that if the cost per unit is lower it will mean our overall cost will be lower too. We have with this believe another, implicit, assumption – that our effectiveness (making money) will not be affected negatively by this choice, or that at least, the negative effect it will suffer is not high enough to justify the additional cost of not batching.

There are many areas in life we batch, and to further narrow down the discussion, let’s review batching in supply chain environment and try to answer the following question – Do we really get the lowest cost and is our choice really not negatively affecting our effectiveness (or alternatively the negative effect on effectiveness is smaller than the additional cost of not batching)? OK, let’s check!

Where do we batch in supply chain environments? Well, where not? We batch when we buy as the seller often time will offer a discount (price per unit), we batch our shipments, when we sell (both when we ship from suppliers as well when we ship to our consumption locations), as the cost per unit of transportation is lower, we batch our warehouse picking, as the cost per picked unit is lower, we batch handling shipping documentation and invoices, as the cost per processed document is lower (we call it efficiency). And let’s stop here. (However, there are many other areas we batch that for the time being we will leave out. Still, I urge you to challenge the validity of the cost effectiveness assumption. )

Before we review each one of the above mentioned batching practices, please allow me to review the mathematics behind the batching “science’ (quotes again, soon it will be clear why). The following graph is the common way cost is related to quantity:

The basic ingredients are an ascending cost (red arrow) that grows as quantities grow (like inventory handling cost), a descending cost (green arrow) that shrinks as quantities grow (such as transportation cost) and their sum (the blue line), which mathematically is curved and has an optimum. How lucky we are. The calculation for the lucky optimal number (called EOQ – Economical Order Quantity), is done using the following equation:

If you review this beautiful mathematical equation you will see its components are:

•         Annual usage in units – which is mostly an estimate
•          Order cost – which is an estimate (and suspiciously related to the estimated annual consumption and the number of orders we are trying to calculate)
•          Annual carrying cost per unit – which is again an estimate, and as suspicious as the previous number

However, the beauty of a mathematical equation is that it results with a very accurate number. So, we take estimates, guesses and highly dependent numbers and drive out of them a perfectly accurate number. Strange, isn’t it? Please note one more, extremely disturbing fact about this equation (and unfortunately this fact is prevalent across most inventory management practices) – this equation totally ignores the element of supply time, which is by far the most important element in any such decision.

In case you do not use EOQ (or any of its similarly dubious derivatives) the relevancy of batching still calls for a review.

As the perceived savings are resulting from similar considerations in all cases, we will start with reviewing the cost savings. What are the cost savings we believe we achieve by batching?

-          Purchasing price per unit, which often time actually is lower for larger quantities.

o   However, often batching leads to buying larger quantities vs. the quantities required within the time it takes the supply point to deliver to the consumption point. In these cases obviously the total cost goes up, as even though we pay less per unit, we pay more for the full batch.

§  Example:

·         If a supplier can deliver once a week, and the minimum batch is sufficient for one month, the minimum order quantity is 1,000 units and price per unit if we buy 250 (which is what we need) is $10 and the price per unit if we buy 1,000 (the “economical” quantity) is $9. Then we pay $9,000 instead of just $2,500. In my mathematics $9,000 is greater than $2,500, does anyone get it otherwise?

-          Handling cost per unit (Transportation, Storing and warehouse operations, Administrative activities such as: shipping documentation, orders, invoices, etc. and Overhead(

o   Handling cost per unit is a calculation that allocates fixed cost of handling to units. We do not really know this allocated cost before a period ends (as only then we know how many units we actually handled). More than that, this cost does not change with a change of unit, but more with leaps in units. You do not really lay off people, or recruit them just because you have one more unit to handle (at lease for most of the cases). So, often times the savings are fictitious, as many CFO’s know when they ask – why can’t we see these saving in our bottom line?

o   Obviously when we have more we need more space to store, more people to handle, we have more obsolescence and scrap, we produce more, and we ship higher quantities. All of these could not have a reduction effect on cost, do they?

Naturally evaluating the effect on revenue generation is more complex as unlike cost, it cannot be calculated, it can only be estimated (read again the difference between cost and revenues explained in the beginning of this article). Therefore one should ask himself what is the probability for the effect on revenues. Accordingly, the next part is providing a glimpse into the lost revenue generation opportunities, to their probable effect, and their probability of materializing. And clearly, they do not always apply. Let’s review the effects of batching on effectiveness:

-          Batching when we buy

o   When space is occupied with inventory we do not need, we cannot hold inventory we do need (for example in stores). Thus we are selling less.

o   As inventory and time are synonymous, and forecast accuracy deteriorates with time, the higher the inventory the more it is liable to misalignment with demand and more occasions of shortages happen.

o   As long as some items have large quantities of inventory, replacing them with others becomes an issue delaying the generation of revenues from alternatives (new products).

o   Example:

§  A pharma retail chain. The facts that help estimate probability:

·         53% of SKU’s have inventory higher than their maximum level in the stores.

·         This excess inventory occupied more than 30% of the stores shelves.

§  The effect of eliminating batching

·         The overall cost went down by 6% (even though transportation cost went up by 2%) and the revenues grew by 15%, resulting with increased profitability of about 10%.

-          Batching when we ship/ sell

o   When we have Minimum Order Quantity (MOQ) we, at times, incentivize  our customers to postpone buying (or not buy at all)

o   When our customers buy quantities larger than they need, they are not likely to buy again until they use the quantity purchased resulting in delays in our ability to introduce new products to the market

o   Example:

§  A Packed Consumer Goods producer. The facts that help estimate probability:

·         About 25% of the retail chain stores the company is selling to, do not hold the company’s products (and obviously the company has zero sales in these stores, even though the chain as a whole is a customer)

·         The company has a MOQ policy that may represent to some stores inventory much higher than their maximum targets

§  The effect of eliminating batching

·         The overall cost of shipping and producing went down by 4% and the revenues grew by 8%, resulting with increased profitability of about 8%.

-          Batching warehouse operations

o   Batching warehouse operations means the warehouse refuses to service orders smaller than the MOQ, thus these sales are either lost, or delayed.

o   Example:

§  A central warehouse of a large retail chain. The facts that help estimate probability:

·         About 20% of the SKU’s in the stores have no inventory while there is sufficient inventory in the warehouse.

§  The effect of eliminating batching

·         The overall cost of warehouse operations went down by 5% and the chain revenues grew by 8%, resulting with increased profitability of about 3%.

-          Batching documentation handling

o   Similar to the warehouse operations, batching orders means the orders will not be released when required, but rather when there enough of them (or when the fixed time for that arrives). This means that sales that required the purchased items are delayed until the processing of the documentation takes place, or lost.

o   And, for received orders they will not be processed until such time there are enough of them (or when the fixed time for that arrives). This means that even if inventory is in place it cannot be used for its purposes as it is not available for that. Obviously, it delays or even leads to lost opportunities.

o   Example 1:

§  Maintenance parts distributor. The facts that help estimate probability:

·         With a policy of ordering from suppliers monthly, about 24% of the SKU’s are either not available at all, or insufficient for their demand.

§  The effect of eliminating batching

·         Availability went up to 99% plus, cost stayed the same, and sales grew by 15% leading to 6% improved profitability

o   Example 2:

§  OEM producer of sub-assemblies for the defense industry. The facts that help estimate probability:

·         With a policy of processing the receiving of purchased items once every two weeks, about 20% of the orders are supplied in a delay even when the purchased items are in the plane warehouse.

·         The Plant pays about 2% of its revenues as late delivery penalties.

§  The effect of eliminating batching

·         Delays went down to less than 5% of the orders, and penalties to less than 0.5% of the revenues.

All of these examples are demonstrating one crucial point – batching, more often than not, does not result with the aspired cost savings and definitely results with the loss of the hoped for revenues. It is not that the batching practice is insensible; it is the policy that is. It should not be used automatically, and should be reviewed on a per case basis, BUT through thorough evaluation of its full effect on the total cost and the revenue generation (and not based on dubious measures that may, or may not indicate it, such as cost per unit).

Batching has yet another meaningful and harmful effect, it damages flow and productivity. These effects are also falling into the category of “difficult to quantify” but still are so bad, that the entire Japanese philosophy is based on the elimination of flow disturbances and is commonly called “one piece flow”. Flow is of utmost importance to speed, reliability, productivity and efficiency of any operational system and batches are negatively affecting these parameters as they contribute to elongation of lead times, to increased obscurity, increased cost, increased obsolescence and reduced effectiveness, and all of these inevitably lead to increased cost and reduced effectiveness.

Do not be afraid to assume effect on revenues, we do this with any new product we develop, with any new market we enter too, with any price decision we make, with any promotion we choose. And mostly, all of these are riskier, costlier and far less probable to affect revenue generation vs. the elimination of batches in the appropriate places.

Batches are a meaningful, mean and well-disguised enemy of productivity, sales, service and, surprisingly of cost too, so go to war, you can only win!