The Purchasing Problem

The commercial relationship between a customer and a supplier is a zero-sum game. When both negotiate about prices, the customers gain (= better price) is the supplier’s loss (=lower profit). This is the bazaar model.

On the other hand, there is a win-win part of the customer / supplier relationship: This is about improved products / services that provide greater customer value and higher supplier profits.

The problem for the customer and supplier is, how to move from the zero-sum price negotiations side to the win-win with higher customer value side?

Of course, the supplier arguments about higher customer value of its products. But can the client trust the sales guy? Can the client proof the sales guy? Does the client really needs this kind of value? How is the higher customer value priced by others? What might be a fair price for the level of value provided? Questions over questions…

The rest of the text explains concepts and methods for production material. That means things like plastic parts, machining parts, etc. The concepts aren’t limited to parts. The concepts are useable for services and of all kind of parts, modules, etc.

Current Approaches

To answer some of these questions clients today invest high efforts using different approaches. Let’s take a look, what approaches are there and what kind of problems these have.

I ordered the concepts from “bad” to “better” by purely subjective opinion.

Benchmarking

A lot of purchasing departments like benchmarking. The idea is, to take a look what pricing others dealing with similar parts have. Or in other words If I know what someone else paid of a specific part, I can see how good I’m doing.

The idea is easy and striking. However, doing it is mostly impossible. Most of the time consulting companies are offering “Benchmarking Projects”. Arguments and setups are mostly always like:

“We have a database of parts and prices. We collected this from several projects and the experience we have done such projects for years. We will take your purchasing data and match it with our benchmark database and then, we can tell you how good your pricing is.”

Ok, granted. Let’s ask a couple of questions:

• Actuality of Data: How old is the benchmark data you have? How can we be sure that the price information is actual? How do you maintain the data? What sources do you use in order to update your price information?
• Match of Data: How do you match our parts to your parts? How are differences (quality requirements, specification differences, etc.) in parts are handled?
• Match of Peers: How are differences in sectors are handled? How are differences in countries are dealt with?

And so on, and so on. It all comes down to one killer question:

“How do you make sure you are not comparing Apples and Oranges?”

I think, one can’t. Benchmarking for pricing is simply not possible in most cases. For highly standardized products and services, it’s possible. In this case, the market is normally highly transparent anyway. Hence, you don’t need benchmarking in this case.

Cost Structure Analysis

Wouldn’t it be nice to understand how the supplier came to the price he is offering? If I can follow all steps, then the pricing is correct & good.

That’s the idea of a cost structure analysis in a nutshell. Yes, that approach has a few appeal.

Executing a cost structure analysis is quite a lot of work. One is required to fully understand how the supplier operates. So, we have to visit the supplier, take a look at the shop floor, the production steps, the company structure, the financial situation etc.

Because we need such a lot of internal information from the supplier, it’s always a balancing act for both parties. The supplier doesn’t want / can’t give us all information. We are always suspicious on everything we get.

At the end of the day, one tries to model the supplier’s company with all aspects as good and close to reality as possible. Based on this model, a part price is calculated, which is taken as the “more realistic” price than then one offered so far.

Of course there a lot of problems associated with this approach too:

• Estimation Errors: If one doesn’t get all necessary information, things have to be estimated. Can we tell how sensible such an estimation error impacts the calculated price?
• Calculation Schemes: It’s most likely that the supplier uses a different calculation scheme, then your “cost structure analysis” tool. There is not any right or wrong how to do price calculations. Which one is correct?
• Ignoring Facts: No “cost structure analysis” tool can handle all facts. A supplier might follow a strategy that will yield better products in the future, you can profit of. Nonetheless, you are not taking this into account. A strategic value, can’t be coped with. How do you take this into consideration? Just ignore it?

Again, I could continue and list more problems. The result of such an analysis has a pseudo accuracy. It looks magic. It’s close enough so it could be true. But it has to many problems.

This way narrows down the limits of a bazaar negotiation for both parties giving them the illusion of a some how systematic evidence.

Shadow Calculation / Bottom Up Calculation

This approach shares a lot of the properties of a “cost structure analysis”.

The main difference is, that the approach takes much more technical and production related parameters into account. And, it uses the expertise of “experts” to build up a company and a production model that is “ideal” in their view.

So, we don’t care how the supplier operates at the moment, but are creating a model supplier, that would be a perfect match for what we want to buy.

There are, again as you can guess, some problems with such an approach:

• Correctness: Who is right how to produce a part? The supplier or the expert? How effective does the model match reality?
• Realization: Let’s assume the created model is correct. What do you do with the consequence if you can’t find a supplier that matches your model? How much of a price difference are you willing to pay more?
• Effort: Building up such a bottom up calculation model is full of work. How you keep it actual?

And so on. The simplest answer for a supplier in such a case is: “Well, we understand, but we are not like you modeled it. We can work in this direction here and there, but that won’t gain as much as you have calculated.”

Part Comparison

All three approaches so far, share one property:

They don’t put the parts into the center of action.

We are interested in finding out if the pricing for our part is appropriate or not. Nothing more, nothing less. So, why not “somehow” compare the parts we have to see how these are related?

Yes, that makes sense. But how?

Well, we can compare similar parts. We can look at some properties of these parts and see who is doing best for which property. We could normalize a property like:

1mm costs X, than 20mm cost 20 * X

The idea sounds very good, but has a couple of problems as well:

• How to deal with more than one property? What if you find out that the price of part A is best, when looking at millimeters (length) and the price of part B if you look at grams (weight)? What price to take now? Calculate an average?
• How to handle the law of “economic of scale” or “diminishing returns” which state that things are not linearly related?
• Which properties to consider? How do I know if these are relevant?

Conclusion

All ways to proof that a pricing is good and fair lack important properties and hence are not really suited. Nevertheless they are used because there are no other approaches known.

Well, it looks like that approaches which take the actual parts that are purchased as the sole information source offer some advantages.

We will look at how this can be done in the next posting.