Choosing reliable subcontractors

How do you choose a good Board Loader? Well, that's a hard task. Choosing a good PCB laminate manufacturer is similarly hard. In reality, you never know if they are good or bad until they "stuff up" and you need to go find a better supplier. Most suppliers say the right things but "do" something completely different, sometime during the relationship.

Choosing a subcontractor is made more difficult because we all have slightly different objectives - one supplier may be a better fit for company A, another a better fit for company B. Add to that, a suppliers reaction to problems (one of the best tests, in my opinion) is going to be modified by what they think of you as a client. But we have to have some ground rules, there has to be some logic to it, or otherwise the selection is not worth doing.

It really helps in evaluating subcontractors to have a solid knowledge of their trade. I can't tell you nearly enough on these web pages - either read up on the subject yourself or employ someone who knows.

Both PCB assembly and PCB etching are broadly similar businesses in a way. They both have to be large to make money. They both critically rely on Systems, Equipment and People, in that order. A failure of the equipment or people is supposed to be caught by the systems. There is a large capital investment in Equipment, often helped by ingenuity on the part of key employees (for instance the owner or software wizard). And their single biggest problem is people - when people don't follow the systems, or just don't turn up for work. Their next biggest problem is equipment, but that is often tied in with the large capital costs - either it is old, or they built it themselves.

Don't respond to spam. This is very much opinion, but I don't think answering undirected email is a good way to find a supplier. For the past few years (and perhaps because we also deflect marketing phone calls) email lists seem to be the way offshore companies have chosen to "market" to us. I don't think it is healthy, the few undirected email contacts that I have followed have not resulted in successful supplier relationships. It is quite obvious if the email is impersonally directed to an old email address that the sender is mailing thousands of others. Make it a little harder for them to get your business. Check if the email is personal

Don't factor reusing tooling costs. So often changes are required between runs anyway, and it seems it is not worth staying with a supplier just to avoid paying again for tooling. The reason to stay with a supplier is because they did a good job last time. We avoid ETest on small (ie prototype) runs of boards, it saves us a few dollars, and if the manufacturer is good there are rarely problems.

Don't just order Australian quantities. Years ago boards purchased in this country were pricey. You may be accustomed to ordering a small run to start, of say, 25 PCBs, perhaps each of 75mm x 140mm for example. That would be reasonable, locally. But offshore the market is competitively pricing double sided work down to about $80 per square meter - that order was only 0.25m2: $20 worth of boards, but at least $120 worth of setup and shipping. You should probably order 50-100 pcs of that board, just on the off chance you may need the larger quantity.

Don't order around Lunar New Year. Everyone goes home and forgets how to do the job. The quality drops, and your order gets lost. You can say the same about Christmas over here, ofcourse.

A visit counts for a lot

Whether it is a PCB laminate supplier, or a Board stuffer, your subcontractor is an important, expensive, critical part of your supply process. You are forming a relationship with them. People form relationships best face to face, then second best when they can see each other (we don't "skype" much, but we imagine this is the case), then third best when they talk by phone, then the rest. A visit helps a relationship form, and persist. If you are visiting their production facitilies, that is a double bonus - you may well learn an incredible amount about their culture, attitudes and processes. How much did they show you? Who did you talk to? That is the total value of the visit. You should make up your own mind about who you want to talk to, but my pick is it be both technical people and management.

When you have the face-to-face at a trade show you do get a chance to make a better relationship - but realise they didn't get a chance to make a relationship with you. Working at a trade show is pretty numbing - there is a constant stream of people, at the end of it you just have a long list of names, each one with a question beside it (Can you deliver multilayer before the end of the week - We want reliable Flexi-rigid, but requirements 1m2 only - Can you beat CTechs price on this board) but you don't remember a single face except your tired co-worker.

When someone visits you at your premises, (and same at a trade show) remember they are either middle managment or more likely a sales person. That is not necessarily a bad thing, you just need to temper your analysis. It costs you very little in effort to meet with a sales person, and if you have spent a few minutes in preparation you could give them a useful task - a competitive quote, technical problem, or set up a special order. They are getting a chance to form a relationship with you, but you are getting less of a chance to evaluate their business, in a way.

Problems with quoting, communication, or delivery.

This is silly! Move quickly, is there a problem with something you are asking for: Seriously, have you put a chip on the board that they can't find, is the information in a format they can't read, are they expecting a purchase order in advance of the work? Otherwise, find three other supplier candidates right away, before your project gets paralysed. If a supplier is being asked for a fixed-price development quote, that can take quite a while - but for assembling a PCB? The quote should take just days. Problems with delivery just should not happen with online tracking, if they do you have to question whether your sales contact is actually making the product or selling your work on to others.

A few times in my career I have been in the situation of waiting too long for a quote, (as in more than two weeks, when we really needed it more quickly) and every time it has transpired that the supplier was no good for other reasons, you need to set yourself a deadline and move on.

Why so much emphasis on problems?

There is a best fit subcontractor for every company. When we are looking for subcontractors, we put a scores on technical competence, information and cost. Perhaps every company puts a fair score on cost, but I am not certain. We can't judge the other factors - quality, and consistency - until later. When everything goes well, the relationship between the supplier and client develops, but is hardly tested - it is when things go wrong that it needs attention. That is why we are focussing on problems - it is easy when the order goes out, the product just arrives, and everything meets spec - there are no problems, it needs no attention.

Supplier: We'll give you work

Untrue. We do design work, our subcontractors make the boards. But most board assemblers are large companies with a big capital investment, and they have their own design departments. They are in active competition for the end client. They will take the first opportunity to blame the designer for a manufacturing disaster, and offer their superior integrated design and manufacturing service. However in my experience the client just ends up with a bad taste in their mouth and walks away from that board assembler, anyway.

Supplier: Let's cooperate

Debatable. Board assemblers, being large companies, frequently DO have a few local contacts that can assist a company in penetrating the market. However, a quote from a friend of mine, Fransisco Shi: "The chinese version of cooperate is 'You buy from me'." It is uncertain that a company would help you long term to get a foothold in the local market, they would be more interested in activities that further their own goals directly.

Supplier: We can help you design your product

Absolutely true, but lost in translation. The design may well be faster. The design work will probably be mediocre, but that won't really matter as what you need is the final result - it does not matter quite so much if it uses a bit more power, or has a rather odd user interface. The design will be flexible to a degree - when you give them feedback straight after receiving the prototype they will redo it and resubmit (but don't wait too long). And the work will be quite cheap - but this is the point - it's not really cheap at all, because you will never own it. You have instructed an overseas manufacturer on how to make your new product, and they have as much respect for intellectual property as you have for their culture. Maybe less.

Client: We will have each supplier make one part, and integrate here

I think that works. But I still think that if you are having electronics designed overseas, then you lose control (and ownership) of your product. The electronics and firmware form the core of your design. We frequently overestimate how much work is involved in remaking a design - sometimes the key to an offshore competitor is just to be shown HOW to make it - what features the product needs. Ofcourse it will cost more, but if you want to make and sell a product, I think that integrating the final steps here is what will work. You will now also be looking for a different class of supplier - board stuffers and PCB manufacturers, rather than product manufacturers.

A suppliers reaction to problems

When there is a problem, what happens? That is one of the best indications of the fit between PCB manufacturer / Board Stuffer and us poor sods trying to turn out good product.

When the supplier contacts you to tell you there is a problem, you need to score them negative points. The better suppliers have systems so the problems don't happen often. The supplier is assembling a circuit board, not launching a satellite. If you have Suppliers that are always contacting you about problems, it will occupy your whole attention and be a miserable way to subcontract. Next: When did they call you? Problems don't happen the day after the goods were supposed to be delivered. What is the suppliers tone, attitude, and response? This tells you volumes about the way the company works internally.

Being pragmatic we have left out "the cause of the problem". The truth is, 66% of the problems will be labour, the people on the ground, 33% of the problems will be equipment. In any decent compnay, none of the problems will be materials supply or systems and procedures. The systems and procedures should have caught any errors by people, any equipment malfunction, and any problem with the information you supplied. Whatever they tell you was the cause of the problem may or may not be fully accurate - you could score them points for honesty.
Problems can only be fixed by people, the response to a problem is your most valuable indicator of the value of those people, and their ability and independence.

When you find a problem after delivery

You need to accurately paramatize and document the problem. Clearly and concisely show what the problem is, how it differs from samples or instructions, and what percentage of units are affected. Photographs help. If there is an obvious cause, draw that conclusion at the end of your report. Get a basic idea of how it will affect your delivery schedules, but it is really open as to how much of this you should tell your supplier. Send the report to your supplier. Then start working through how it really impacts your delivery - is there a workaround? How much slack time is in the delivery schedule? How does it impact you? This really determines your actions - don't lose a client because you have a useless PCB manufacturer. Most of our clients are pretty understanding, but there is a limit as to how many times you can pass on your subcontractors excuses.

You should expect a prompt response to your report. There will be timezone diffences, but if there was a substantial problem you should expect a pretty much immediate: "Sorry that there are problems, thank you for your information, we are investigating". The investigation and drafting a response may take a day, often it involves someone digging out samples of your boards, checking them, and interviewing the person who stuffed up your job. (Okay, a touch of cynicism - but this happens pretty much all the time)

How do they offer to compensate

The best you can hope for from a PCB manufacturer is to remake the boards in perhaps half the time that they had promised before. There is not enough money in PCB manufacture to pay for the boards you have assembled that do not work, your wasted components and time. The best you can hope for from a contract assembler is to rework the boards as quickly as possible - and maybe to pay freight both ways. If you take the obvious course of reworking the boards yourself to get the problem solved more quickly, the PCB assembler probably won't assist much at all.

A supplier offering to compensate 'on your next order' is really an empty promise. It could also indicate someone who has no choice - such as a sales agent acting as middleman. You want to avoid sales agents if you can, because while they seem to make communication easier (that is their skill) they actually make resolving problems harder, because it is all third hand. But, you may as well reply immediately asking for details in writing.

When a suppliers reaction to your report is not to compensate, but to question, you really should answer the questions but probably only as an afterthought - they are, in a way, making their position clear. We had one PCB manufacturer that sent us a delivery of boards that had been substantially hand reworked. There had obviously been a production problem, and their solution had been to grade the boards as best they could and patch up track problems with a soldering iron, and soldermask flaws by hand with UV cure paint. We sent a report and photos, they came back with questions - what zoom had we taken the photos on, etc? You have to draw the line somewhere and say, even if they DO offer to remake, would we use them again?

The blame game

A good subcontract assembler or PCB manufacturer should not make faulty product, even if you tell them to do it. They should have systems and procedures in place to detect problems and stop them before they end up in your loading dock. While you might think that the failure on your PCBs, out of all possible failures, could never have happened before - be assured it probably has happened many times, just not to you. Some subcontractors do try to blame others, especially the designer. And, of course, there is always a grain of truth behind the story, or else it wouldn't sell. But see it for what it is - avoiding responsibility.

In theory, there is no difference between theory and practice. But in practice, there is. -- Yogi Berra (attributed to) :-)

©2013 AirBorn - Last updated 27 April 2013