Formula compare – production efficiency

When we develop a new product, we calculate the estimated cost. We will then put the new product into production. However, if it is not a drug, then the amount of material we actually put into the product might differ from what we’d planned. We might even substitute materials with different costs.

Our production goal is to create a good product within the stated spec. The cost of the materials required to produce a batch might differ from the cost we’d come up with within the calculation.

Example 1: We produce chocolate. We develop a product composed of 60% cocoa beans. To achieve this 60% cocoa, we test the product with a 62% +/- 2% tolerance. (The product must NEVER be under 60%.) The other ingredients are very cheap.

We will call them “other ingredients.” When we calculated the planned cost, we did so by using the following formula:

Cost of raw materials = 62% * Cost of cocoa beans  + 38% * Cost of other ingredients

Now let’s assume that cocoa beans’ cost is $2.20 per kg and that the cost of “other ingredients” is $0.30 per kg.

The cost of the materials for the product will be $1.48.

Assume that the actual production is done using 64% cocoa beans (still in spec.). The actual cost of materials for this batch will be $1.52! The cost of the product just climbed by 2.5% without considering it during the costing.

The problem worsens when we think that we market the chocolate as 60% cocoa when we actually use 64% cocoa beans. Let’s look at the following table to understand the difference:

Now let’s assume that we have a product that we sell for a 7% profit. Can we afford to produce it if the raw materials’ cost is 5% more than we assumed?

We can easily erode our profit by 20%-50%

Example 2: Let’s assume that we planned to use 62% cocoa beans, and we actually put 62% in the product. However, the regular cocoa beans were not available, so we used expensive cocoa beans. We might have a better-quality product, but the price for the customer did not change. So, if we use 10% more costly beans, we might increase our product’s cost by 9%!

We need to answer the following questions:

 1. How can we identify a significant change between the actual production and the planned formula?

 2. What should we do when we identify a significant difference?

Identify the abnormal change in production cost.

“Formula compare” meeting

Each week, we have many batches running on many tools. Every batch might produce different products and different formulas or even use different “similar” raw materials, which means different serial numbers for the raw material.

So, how do we know which batch to focus upon?

We look at all the batches created during the time period, and then we sort them according to the difference between the planned material cost of the batch and the actual cost of the materials used. Next, we sort the batches according to the difference we found. We will start with the batches that will have the most significant difference in terms of absolute value. (If the difference is negative, we used more expensive materials; if the difference is positive, we used cheaper materials, and we should see how we did it.) We are looking at the differences both by percentage and by absolute value. Sometimes we will have a high percentage, but almost no absolute cost, and we cannot do much with it. Sometimes we will have a small change in percentage but, because of the batch size, there will be significant savings or waste in absolute value. We flag the batches that we want to check. This entire process can work almost automatically. (I will provide the algorithm in a few paragraphs.)

The formula compare team owner will review the flagged batches to see if they make sense (i.e., no garbage data).

The owner will organize a meeting with the best people from R&D, production, process, and procurement. Why the best? Because we want to identify the problem, make decisions, and follow the execution. The meeting will last up to 1.5 hours. (Don’t do longer sessions, as you will lose focus.) During the meeting, we will see the flagged batches, as well as what was planned and what actually happened. We will start with the batches that experienced the most significant changes (whether plus or minus) in terms of cost.

The meeting participants will get the following tables for each batch and will try to understand why the batch was produced that way. In the following case (in the picture), we can see that the production used a cheaper material but managed to obtain the same quality. That saved $7,000 for the batch. Now, we determine whether we can produce this product and other products using the cheaper material as the standard. (In this case, we achieve a savings of hundreds of dollars by doing almost nothing. This is a true case.)

At the end of each meeting, we will write down all the action items and the owner of each item. When we start the next meeting, we will review all the unclosed action items. For each action item for which we did a change, we will show the yearly projected savings so that people will see the importance of the meeting.

How to construct the data table

1. We define a standard unit cost for each raw material. This can be done in one of the following ways:
   1. Moving the average of the unit cost in the last X orders.
   2. The unit price cost from the last order.
   3. We enter a manual cost because we know that prices might increase or decrease soon.
2. Show the planned formula quantities as a percentage of the actual quantity being produced. For example, if we produce a 1,000 kg batch and we have a formula with 10% of raw material X, then we will see “100KG” next to the X material in the table.
3. Use only the standard unit cost and not the planned unit cost generated during the R&D phase. We need to compare apples to apples.
4. If we rework some of the materials, we must adjust the data accordingly.
5. We must put both tables next to each other and see each material and its description, the percentage in the formula (in the planned table) or the actual percentage (in the actual table), and, of course, the unit price and the total cost of the material.
6. At the bottom, we put a price per kg, the total cost of the planned formula, and the actual formula.
7. In the header, we put the finished good product name and number, the tool used, and the date and time when the batch started.
8. We should be able to see the data for every batch we pick.
9. We calculate the difference between the actual cost and the planned cost of each batch. We put the difference in absolute value (because we want to deal with over-cost formulas and under-cost formulas, as both can teach us a lot). Then we sort from the most significant number to the lowest number.
10. Upfront, we decide on the minimum difference in cost that we want to deal with. If no batch matches the minimum, we cancel the meeting.
11. As stated before, for every change we make, we should calculate the yearly saving so that we will always know the value of the meeting.

 

What if we find a problem?

The following are real-life scenarios that came out of the meeting and were implemented in many cases.

1. 1. There might be a shortage of cheaper material because we did not estimate the correct amount that we needed or because we had not run the MRP program in a while, so we did not fix the problem sooner. In that case, it is essential to calculate the amount of damage. Usually, more than one batch will use expensive material. Once we know the cost of using the expensive material, we can decide to increase the safety stock, improve our sales prediction, or run an MRP more often. It is easier to decide whether to increase the safety stock when you have the real cost involved in using the more expensive material 🙂
2. This scenario is the opposite of the previous scenario. We used an expensive material instead of a planned cheaper material. Reason for that:
   1. There might be a shortage of the cheaper material, and it could be because we did not estimate the right amount that we needed, or we ran the MRP program once in a long time, so we did not fix the problem sooner. It is essential to calculate the amount of damage, usually, there will be more than one batch that uses the expensive material. Once we know the cost of using the expensive material, we can decide to increase the safety stock, improve our sales prediction, or run an MRP more often. It is easier to choose to increase safety stock when you have the real cost of using the more expensive material 🙂
   2. We might find that the material usually arrives late and, thus, miss the due date. In this case, we will want to use our procurement department to measure the supplier’s delays in all the products it supplies. Then we can educate the supplier so that it reaches the standard we require. If that doesn’t work, we can increase our lead time, sending our purchase order sooner.
   3. Another scenario we might encounter is that the more expensive the material is, it is actually better for production. Perhaps it results in less waste, so production wants to use it instead of the cheaper material. If this is the case, we should calculate the savings we achieve when we produce less waste than the extra cost of the material. If the reduction in waste makes it worthwhile, then we will need to change the formula. If not, we must explain this to the people working the machines and then define the expensive material as NOT a substitute for the cheaper material.
   4. Sometimes the standard cost has not been updated correctly. In this case, when the cheaper and expensive materials swap places, we don’t always see it at the standard cost. In that case, update the standard cost.
3. Finding opportunities to change a symmetric tolerance to an asymmetric tolerance. I will explain by using an example: A product is made of 20% polymer and 80% talc. Talc costs $0.50 per kg, and polymer costs about $5 per kg. The product specification states that the product consists of 80% talc +/- 2%. (This is basically a standard product in the market.) So, if we use 78% talc, we get a positive result from the lab. However, we used 22% polymer instead of 20%, which increased the cost by 6.5%. If this is an industry of 7%-10% profit, then we almost didn’t profit from that batch. After we intervened, we changed the test specification to be 80% talc +2% and -0%. (Notice that we did not change the marketing specification of 80 +/- 2%; we changed only the test specification.) This ensures that we will never use less than 80% when we produce this product, and we will never lose without noticing it. This change saves millions of dollars because we did it to every formula with a huge amount of talc. The cost of the change: ZERO dollars.
4. Do what we did in section 3 for every formula in which there is a huge difference in the cost of the most expensive and the least expensive materials.

Summery

Often, we can save a lot of money and improve the production process by better planning our products and the test specification. Experience has taught me that, among the improvements we make, the ones that are difficult to spot are the ones that create the most benefits and give us an edge over our competitors. The reasons for that are:

1. Easy to do: It is straightforward to change a formula or a test specification (unless you produce medicines). We can sometimes involve the customer and share the benefits.
2. Consistency: When we spot a place to improve one formula, we can usually improve other formulas and use the same principles.
3. R&D change of mindset: When we learn the importance of asymmetric tolerance, we can apply it to every new product that we develop.

The downside: You must form a great team and take time away from your most valuable people.

As always, I would love it if you clicked the like button on Facebook so that I’ll know whether you liked this post. If you do, I’ll be sure to write more.

Any remarks or comments feel free to send to:

ThePlanningMaster@gmail.com