|
|
|
09/20/2011 03:05 PM
|
|
|
Hi all,
Do you have any comment or tips about applying Double sourcing or parntership with
single suppliers. How will it depends on situition?
|
|
|
|
|
09/22/2011 03:42 PM
|
|
|
Hi Mark,
Just a couple of comments.
1) Sourcing of parts from suppliers is all about risk/ reward
Reward:
- the fewer suppliers you have, the more orders you can place on them
- the larger the volume of orders - the lower the price
- so why might a single supplier not be the most appropriate
Risk (of having a single supplier)
- what if they have a quality issue
- what if they can't supply when you want the goods
- if qualification of a new supplier is not straight forward does the supplier
hold the upper hand in this relationship (especially if you represent a small
percentage of their business)
So, maybe two main suppliers in optimal. The first having 75% of your business
and the second 25%.
2) Re partnerships - one company I worked for used a massive amount of acid that
caused real logistic issues given the relatively small size containers in which the acid was delivered. The company and supplier shared the cost of constructing a large
container in which to store the acid. The supplier continued to own the acid until is was used. The company piad for what had been used on a monthly basis.
Perhaps you have the opportunity to have a supplier use your warehouse as 'theirs'
Cheers,
Martyn
|
|
|
|
|
09/23/2011 11:05 AM
|
|
|
HI Martyn,
Thanks for your Reward/risk comments, i am still confused how to manage supply base, if we use 75/25 principle, we will have a large base, which is not so lean.
COuld you tell more about your current model to manage supplier base?
|
|
|
|
|
09/23/2011 11:05 AM
|
|
|
Mark,
Martyn makes some good comments. I would add to this that there is a hidden cost in dual- (or multi-) sourcing: you must over-design your product in order to accommodate the differing quality levels of different suppliers. This always adds cost to the product, and it's generally a variable cost (i.e. it scales with the number of items produced by your suppliers).
As a simple example, suppose that you source metal rods that have to be a certain diameter, 10 mm +/- 1 mm. This is your specification. Supplier A manufactures these with mean 10.3 and standard deviation 0.18, resulting in a Cpk of 1.33; Supplier B manufactures them with mean 9.9 and standard deviation 0.2, resulting in a Cpk of 1.50.
If the Cpk of 1.33 is an acceptable quality level, we might conclude that everything is fine and both suppliers are supplying good parts. However, it also means that Supplier B is supplying better parts than you need, and this is likely coming at a higher piece cost. Additionally, parts from Supplier A will have higher scrap, rework and customer complaint rates than parts from Supplier B. If you pay the same for parts from both suppliers, then you're paying too much for Supplier A's (inferior) parts.
Either way, your company is losing money by either paying too much for parts from Supplier B or by paying more on scrap and rework for parts from Supplier A.
Worse, these rods are most likely mixed once they are received into the assembly plant (after all, they're considered equivalent parts by intent). If there is a quality escape from one supplier, you will have to quarantine all of the mixed parts from both suppliers (and maybe issue a recall on all of your shipped assemblies), even though only about half of your parts are potentially defective. This will roughly double your costs for correcting any quality issues detected after incoming inspection.
Finally, it is usually easier to shift a process mean than to reduce process variation. This means that will likely be relatively easy and inexpensive to improve Supplier A to a quality level with Cpk near 2.0, but relatively difficult and expensive to achieve the same quality improvement with Supplier B. If quality improvement is needed, Supplier B, who started out looking better, is going to cost you a lot more for the improvement.
For these reasons, Deming always argued that it was better to single-source and ensure the long-term profitability of your single partner than to buy the "same" parts from multiple suppliers. In some cases, though, this relationship is impossible, especially if you only represent a small fraction of your supplier's business and the supplier holds the upper hand (as Martyn points out).
|
|
|
|
|
09/26/2011 10:55 AM
|
|
|
Hi Mark,
It's several years since I worked in manufacturing. I now work in the Finacial sector - which doesn't have the same critical requirements.
However, from memory, a few thoughts:
1) Identify the raw materials that are critical to your company:
a) quality of such items is paramount OR
b) high total spend per annum
(calculate via ABC analysis:
Number purchased per annum * price per item
This allows cheap items that are used alot to be compared against more
expensive items that are used less often.
This also means that equipment spare parts (for example) could be just as
important as raw materials used in your products)
c) likelihood of supply being disrupted (consider what happened following
the Tsunami that recently hit Japan)
For items that are critical to your company, you should consider the nature of the
contracts you have with the suppliers - and treat them as 'partners', not merely
suppliers. That's not to say that suppliers are not important, they are, but if a
supplier is deemed to be a partner then you should be working more closely with
them - letting them be aware of your production plans so that they can prepare
their schedules. Work flexibly together, setup regular audits of their facilities and
process methods. Let your partners visit your facility, show them how their items
are used - they may have suggestions that could help your processes.
Working in this way will necessitate that you reduce the number of company's from
which you buy the same parts - as you won't want to take on the overhead of
activities that a partnership requires.
Cheers,
Martyn
|
|
|
|
|
09/27/2011 03:18 PM
|
|
|
Hi Martyn,
Thanks for your input from a high Financial Level.
I aagree we should do ABC analysis to find critical suppliers, and actually we kow who is important to us.
The problems i often encountered is that we have to invest money for toolings or some certificate at the beginning of our cooperation, when we paid for all of these, they will be the only suppliers in quite a long time. and problems frequently rising during this period.
In case of this ,how can we estimate the likelihood of supply being disrupted , or what we do is the optimonal for saving cost.
In some cases , tools are not universal for every suppliers.
Cheers,
Mark
lintao9166@hotmail.com
|
|
|
|
|
09/27/2011 03:18 PM
|
|
|
I have found this simple idea to be useful. You should have only one supplier per part but two or even more suppliers per commodity. Commodity suppliers should be capable of supplying parts they do not normally supply, but only in extreme circumstances. If a supplier is chronically late, change him out but do not go to another each time. I have dealt with the question, "what if the supplier's factory burns down?", too many times. It happened only once in my 40 years and we were up and running again in two weeks at an alternate commodity supplier. There is no scenario that leads me to multiple part suppliers other than inadequate capacity in any one to meet my needs.
|
|
|
|
|
09/28/2011 09:44 AM
|
|
|
I agree fully with Jerry. One supplier partner for components and two or more for commodities. When I picked up responsibility for a large purchasing organization the standard was three suppliers per part. What did that do? It created a lot of cost and variation due to multiple tooling, follow up, quality systems confirmation, PPAP, etc. and the staff that it took to manage the process. The normal product dimensional and composition variation increased costs in manufacturing. Sometimes the production department would hold one supplier's material to run another because the scrap rate of castings and tooling usage was better. It took five years to approach the one supplier per major component but the quality and cost benefit was work the effort. In ten years we never had an interruption although we monitored potential problems closely.
Ron Turkett
|
|
|
|
>
Community >
Forums
Send to a Friend
