In a previous blog, I discussed some of the myths surrounding the relationship between price and volume. Another often misunderstood subject is “security of supply.” In analyzing our clients' component spending, we frequently see unnecessary costs incurred that create an expensive (and often false) sense of “supply security.”
Lytica's freebenchmarking.com reports reveal the exorbitant dollars in extra spending that many companies are paying for what we term “coding duplications”. These coding duplications have two main causes.
One is component coding assignments where the same manufacturer's part is given multiple client part numbers for any number or reasons, both intentional and through coding errors. The other cause, and one that tends to dominate the extra spending, is intentional sourcing allocations. The question is: Are companies really getting enhanced supply security from their sourcing allocations, and at what cost?
Let's use an example for illustration purposes. A company has decided that it needs two sources of supply for a component on which the annual spending is about $1 million for 1 million pieces. It has two approved sources and decides to split the volume on a 60 percent to 40 percent basis to keep both sources active. The quoted prices it has for this component are $1.00 and $1.10, so it buys 60 percent from the suppler charging $1.00 and the rest from the other. The cost of supply is $1.04 million, of which $40,000 is the added cost for supply security.
My question is: What are you really getting for this $40,000? Are you rewarding the second supplier for its internal inefficiency or the greed that led to it not being able to supply for $1.00?
Clearly, having an engineering-approved alternate source is prudent, but does buying from that alternative add any additional security of supply? I think the answer is “no,” especially if your total purchases from these suppliers is substantial.
The only reason I can think of to run with both suppliers in production is to ensure that no manufacturing changes at either supplier cause a change in your product. $40,000 seems like a high price for this assurance. For low-end components like passives or discretes, no such assurance should be required. For high-end parts, a simple sampling program where a few devices are used each quarter should suffice.
Personally, I like the strategy of having three suppliers qualified for each component, but buying 100 percent of any component's demand from only one supplier. (Note that to be a qualified supplier in my world, you must meet the quality, service, delivery, and price expectations.) With this approach, you buy many different components all for the best price, and, as a result, all three suppliers end up doing business with you at a level that keeps their interest.
Each supplier should have the best price on some of the components. If this doesn't happen, you have chosen the wrong suppliers in the first place. This works well at a factory level but across multiple factories in different regions, you may find that a different qualified supplier has better pricing. If this occurs, use the lower-priced supplier in that region and renegotiate with the first supplier to match its best price in all factories.
Is this practical? I have been impressed many times in my career at how effective supply chain professionals are at solving shortage problems. Many times, the solution involves finding an alternative manufacturer for a part, which requires qualification and rapid re-vectoring of the chain. My proposal involves only the re-vectoring part if any 100 percent allocated supplier failed to deliver.