Some really large factories appeared in the US and in Europe in the early 20th Century. The prime example is the Ford Motor Company and its huge factories. The basic idea was that the higher the number of pieces produced, the lower the unit cost.

This is what 99%+ of Chinese factories still believe. They’d much rather make 100,000 widgets in one batch, rather than 100,000 widgets over three batches.

Why is this a problem? Because this notion is misguided. Toyota started reducing the size of production lots in the 1970s, with great results on costs, quality, and lead times. Since then, leading manufacturers everywhere–including Ford–have decided to process smaller batches.

Focusing on large batches is looking for local optimizations, rather than driving overall costs down. But it also has an effect on quality. Here are the drawbacks of large lot production:

1. The first finished products can appear after weeks of production. Some factories do all the 1st operations, then all the 2nd operations, and so on until the last operation. When do they notice issues? Sometimes at the very end of production, because that’s when they can test finished products…
2. Buyers often ask for “production samples”, or “shipment samples”, to have an idea of what finished products look like. But sometimes the finished products appear very late, So what does the factory do, to send these samples in time? It often “prepares” them individually. As a result, these products are not representative of production.
3. There is a huge gap between engineering (i.e. the development of samples) and bulk production. They are done in different rooms by different operators and sometimes on different machines. The goal is to let the production operators focus on their own productivity, rather than stopping a line. But there are often differences between what a engineer makes in the sample room and what line operators produce. Many issues appear in production, and corrective actions are more complicated to implement at that stage.
4. Factory workers are paid by the number of pieces they make. Sometimes they see they are working on defective goods, but they keep working instead of sending the bad pieces back to the previous operation. I have seen this many times in Chinese factories. Sometimes even the supervisors see it but don’t stop the line!
5. Large buyers can pull much lower prices than small buyers, in some cases below the full manufacturing costs, in part because the factory over-estimates the cost savings it will achieve. After one of two years, the supplier notices that it is in the red and increases prices, causing the big buyer to switch to another maker.

Now try to explain all this to your Chinese suppliers, and good luck… You need to challenge the whole system. To be fair, importers also play this game by trying to fill containers up and improve freight efficiency.

Checking quality in a factory always takes longer than in an importer’s warehouse. Sometimes I give a quotation to a client, who replies “why do you charge 2 days? It would only take me 1 day!”

I never like to be suspected by my clients, so I have to explain the situation to them. Here is what I usually respond:

In perfect conditions it is doable, yes.

If you have to get to the factory, check the quantity, select and open the cartons, check all inspection samples carefully, take the measurements and do all the tests, check all the details of labeling and packaging… It takes time, but in theory it is doable.

The problem is, you might forget to consider three factors:

Less time available for checking quality in the factory

• Factories do not want the inspector to arrive too early (9am is considered early),
• They want us to get out during their lunch break (everyone goes to eat and then have a nap for at least 90min, so the factory is usually closed),
• The last bus is around 6:30pm in the bus station.

Lower productivity while checking quality

• The factory people often discuss the inspector’s method along the way. They are watching the job being done, so no shortcut can be taken.
• Before leaving, the inspector has to call a manager and talk him into signing the list of findings.
• A third-party inspector can never be as familiar with a certain product (and its specs) as its buyer.

The danger of overload

An inspector should never have to hurry up. If he does, certain issues might be overlooked, and the job is not done properly.

He should do his work methodically and follow the procedure all along (i.e. take photos, refer to documents, write down every finding, etc.).

Some clients also try to “negotiate” to get a “better offer”. For example, I tell them we can check 200 samples visually, but we will only check 5 samples for measurements. (This is based on the complexity of the product and the time we estimate it will take.) Then they ask for 10 samples in measurements, hoping to settle for 7 or 8.

This type of discussion usually takes place with new clients, who don’t understand the way we work. They think we are lazy and we don’t try to “go the extra mile” to get their business.

What they don’t understand is that an inspector in a hurry is not a reliable inspector, and they are working against their own interest.

Several articles and books have described the perverse effects of audits for social compliance. An auditor goes to a factory, looks for evidence of non-compliance to local law and some international standards, and produces a report used by importers to eliminate the worst suppliers.

There are several issues with this type of audit:

• Auditors can get much more money from bribery than they get from their “official” employer, so corruption is quite frequent and findings are not always reported in full.
• Auditors are virtually always local employees, since they have to decipher records about working hours and pay, and they have to interview some workers. Using foreigners to reduce chances of bribery would be quite impractical.
• Factories have learned how to fake their records of working hours (and their whole accounting books), and they tell the workers to lie if questioned by auditors. Some “consulting companies” help them if necessary.
• Some buyers actually don’t want the auditors to be too tough; they want to be able to say something like “70% of the factories working for us are 90% compliant”.
• The vast majority of buyers keep giving business to the cheapest suppliers. Audit results are only used to eliminate the most risky factories (i.e. risky for the brand’s image). There is no reward for truly compliant factories, which have to make do with costs higher than those of  their competitors.

I just came upon an interesting report by China Labor Watch (A Case Analysis Of Corruption) which confirms some of the above points. (Note: if you are in China, this website is blocked)

They did a study after Chinese media widely denounced the behavior of the auditors working for one of the largest QC firms (I won’t name them because I believe the 3 or 4 major QC firms all suffer the same issues in China). The story goes like this: the factory supposedly refused to pay bribes, and the auditors came back 3 times–each time producing a failed report. Finally the factory lost orders and went bankrupt.

During [the factory's] first [QC firm's] audit last November, auditors mentioned that an outside consulting company could help the factory to pass its audit. After the failed audit, factory managers received calls from a consulting company that claimed it could guarantee a successful second audit if the factory paid 30,000 RMB ($4,412). The factory refused, and failed the second audit too. Following this second failure, the factory recorded one of the calls with the consulting company, this time with a demand for 70,000 RMB ($10,294) for a guaranteed pass. [QC firm] itself paid for a third audit in which an independent third-party monitor from the International Council of Toy Industries (ICTI) accompanied the auditors. Once more, the factory failed. The factory claimed that in this and a subsequent fourth audit, [QC firm's] auditors took a retaliatory attitude toward the factory, leading the factory to fail and subsequently move to the brink of bankruptcy.

[QC firm] explains that the fraudulent consulting company was not connected to any current employees at the company, but rather was the work of an auditor previously terminated for corruption ([QC firm] has not addressed, however, how the consultant company knew of the supposedly unannounced audit, which suggests that some current employee was involved in this corruption).

China Labor Watch (CLW) sent auditors to factories that had previously been audited, and “of 19 [QC firm's] audits investigated by CLW, 10 or 53% yielded significant evidence of corruption”.

What shape did corruption take? I pasted some excerpts below, but I advise all readers interested in these issues to read the report in full.

Auditors often extort factories to give bribes. The process of extortion begins with a question. During the audit process, after auditors have identified proof of violations at a factory, they will present the proof to factory management and ask how the violation should be handled.

Many factory managers know that when they are asked this question, the auditor is asking for a bribe. Paying this bribe will allow the factory to successfully pass the audit and is therefore in the best interests of the factory (at least, their short term interest). Thus, most managers will respond, “We hope you can help us out, please tell us what we must do to avoid having this violation recorded in the report.”

And if a factory denounces the auditors’ behavior, what happens? The audit fails and the buyer asks for a re-audit, which is not always conducted objectively:

When auditors came to inspect the factory [the first time], they reported a number of violations. One factory manager got into a heated argument with the auditors about the report. This fight soured the relationship between the factory and the auditors, setting the scene for a retaliatory audit.

During the follow up audit three months later, the auditors used an irregular, retaliatory method to conduct the audit. They requested the last 10,000 pages of production records, and announced that working hours were inconsistent with the records. The factory manager once again argued with the auditor, and demanded to be told the auditor’s name and contact information as well as information on the complaint process. The auditor refused to provide this information and simply told the factory to call BV’s office.

What can QC firms do to reduce corruption risks? This is a complex issue, but it starts with small things. I’ll take 2 examples below.

First, QC firms usually force their inspectors/auditors to have a notice given to a factory manager before service starts, with clear instructions for factory complaints. If the auditor’s name is not communicated to the factory, this is already a problem. The inspector/auditor should be clearly identified.

Second, QC firms usually have their own investigation (or “internal audit”) team. It might be effective… But in China, I can very easily picture how such an “independent” body can monetize its position. As written in CLW’s report, “factories fear reporting extortion because they do have some issues and suspect that after reporting they will face an even stricter audit.”

An easy solution is to systematically give re-audits and re-inspections to another QC firm. Yes, I know, they would lose some business, and the practical implications are not very clear. But maybe that’s the price for conducting clean operations!

I often blame the attitude of factory bosses for the quality issues experienced by importers in China. But the problems really come from the business environment. Few manufacturers (in proportion) truly commit to maintaining a quality standard and investing for the long term.
This is exposed very clearly in The China Price, by Alexandra Harney. I am always surprised when a journalist takes the pain to study a subject in depth and uncover insights that are not obvious. This is certainly the case with this book.

Harney explains how Chinese products are so cheap and reports on the consequences on China’s population and environment. I would suggest to read Managing the Dragon (or my review: Why is the “China price” so low?) for a good complement.

Here are a few excerpts about the pressure on prices and the short-sightedness of large buyers:

In the pie of profits made by selling things to people, the manufacturer often gets the smallest slice. Depending on the product, manufacturers can make as little as a few percent or even less.

[Big-box retailers] base their business model on providing goods at affordable prices. Shoppers expect continual price declines, so the retailers, and the middlemen that supply them, demand continual price declines from their suppliers. If one factory can’t provide that, they find another that can.

Despite these pressures, there is no shortage of people in China willing to sign up for a contract to produce for big international brands. The promise of the annual volume of shipment involved [...] provides a powerful incentive to entrepreneurs.

[Factory] managers, many of whom have little of no management training and sometimes no more than a high school education, often accept orders without considering whether they have the capacity to fulfill them.

Here is what a factory owner had to do to get orders from Wal-Mart (and to offer low prices, fast delivery requirements, and a limit on working hours):

“The basic trouble is the workers want to earn more in a short time”, Chan told me. “If any factory has overtime in some area, all the workers will go to it. Workers like overtime. They tell me, ‘I haven’t come here for a holiday. I’m here to earn money.’ If you don’t want to have overtime, they will leave.”

“All Chinese people want more money because it’s the culture… the workers like to work and earn money because they don’t want to stay here very long. After a few years, they’d like to go home and have a very small business”.

Chan asked for help from friends who ran their own factories. They explained how they coped with the divergent demands of their workers and customers: They made one set of time cards for Wal-Mart and kept the real ones elsewhere. They coached their workers to give Wal-Mart the answers it wanted to hear. And they set up factories on the side, factories Wal-Mart never saw but that made its products anyway. In essence, they subcontracted.

Harney goes on to explain that these “shadow factories” usually hire workers who want to work long hours without social security to earn a bit more. Sometimes these workshops are not even declared to the government, to avoid taxes and labor inspections. I wrote before how this type of system is an insult to proper quality management: see How Chinese factory owners treat their workers.

The China Price also describes in great details the factory audits regarding social compliance, and their limits. I couldn’t agree more. Great book.

Factory audits, like product inspections, need to be performed in an objective manner. A buyer that pays for audits of two factories should be able to compare the findings and use them to pick a supplier.

It means auditors should work with the same basic checklist for all their evaluations, and make a few adjustments in function of production processes or client requirements.

For audits of labor conditions, the checklist is usually based on the SA8000 standard. For audits of a quality management system, the most widely followed standard is the ISO9000 series.

I proposed a basic checklist for audits to quality standards (based on the ISO9000 series) on the China Sourcing Blog. Here is the article: Factory Quality System: What Auditors Usually Check.

The idea is to help importers who visit their China suppliers but have no experience in operations/quality management. They can adopt the checklist I proposed as a general guideline to evaluate the reliability of each factory they visit… If they take the time for it!

This morning I was waiting for a bus, and I couldn’t resist taking this photo:

The Chinese policeman is standing by the scanning machine, to make sure all luggages are checked. But his colleague, who watches the screen, is away (he came back 15min later).

It reminds me of some factory audits, where a manager proudly shows you several employees controlling quality in the lines. When one asks to see the records of their controls, the manager usually says something like “I will show it to you later–we have to ask the person in charge, who happens to be away right now”. Of course, in these cases there are no detailed records.

Without records, no accountability, and no data to work with. How can a factory keep operations under control if nothing is written?

People often ask me to tell some “juicy stories” about inspections in Chinese factories. The ones who are familiar with QC expect me to say a lot of things about corruption, but I don’t have any story on this subject. I am a foreigner and it means I don’t have to be “polite”. I never accept to waste even 1 minute having tea with the boss—I’d rather spend that time relaxing in the hotel after the day is over. Discussions are always kept about the business at hand. As a consequence, I have seen very few outright cases of bribery. All I saw were the offers to “have a good time together” or to take that “red pocket” after the New Year. Yes it’s unacceptable behaviour, but it doesn’t make for good stories.

The most interesting inspections, in hindsight, as those where everything went wrong. I had my share of bad experiences, especially when I started doing QC. That’s when I learned a lot about what to do and what not to do.

A good example sticks in my memory. It took place in Wenzhou, a coastal city in Zhejiang province, between Ningbo and Fuzhou. This city is often described as “very entrepreneurial”. More about that later in this post.

I flew there with a colleague, we slept in a hotel close to the factory, and by 9am we had walked our way there. The director, a short and very talkative guy from Jiangxi province, tried to waste our time (sorry, he tried to be polite), but I reminded him we had a lot of work that day. he was really trying to be nice, and we starting thinking there was a rat somewhere. That’s when he told us “we’ll drive you to another factory, where the goods are”. Oh, really? Didn’t he tell us to come to this factory? It was too late to waste time arguing about this, so we agreed to go there with his driver.

What really happened is this: the importer placed orders to a “main factory” that he probably visited and approved. This factory buys the materials and gives the sewing job to smaller workshops, to lower its costs and increase its capacity. There were two different workshops for this order, fortunately in the same city. Since then, I noticed this type of arrangement is very common in China—not only in the garment industry. Uncontrolled subcontracint is probably what people call “entrepreneurial” behavior: doing whatever is necessary to make a quick buck.

Of course, the workshops were late (most of the goods were still in the last operation before packing), and there were many quality problems. I saw the workers pile the garments on the floor, and I wondered how we didn’t find more stains. This is typical of subcontracted work. The small workshop makes a very small margin and tries to go as fast as possible. They don’t even know if they will get the repeat orders, since the “main factory” might give it to another, cheaper workshop next time. And of course, there is no QC supervision—from anybody except my client, but at this point production was finished and it was too late. I already touched on this issue in Why can quality be so bad in China?

After the first inspection was done, we had to go to the second workshop—we wasted more than 3 hours in transportation that day, and we finished (in the factory) after 11pm. I couldn’t decently abort the job, since my client was pushing the factory really hard to ship the goods out as fast as possible. They were already 2 weeks late.

I had a good time when the agent from Hong Kong called me and asked for a report—it was 10pm, China time, and the client was asking for inspection results. When I was talking to the agent, I noticed she didn’t know anything about the situation. That’s how things often work here: people are all too eager to “just sub it out” to suppliers they know, without any follow-up job.

Not only were the 2 workshops doing a poor job, but the other sub-suppliers could also be criticized. The sandblasting on the fabric was too strong and very inconsistent from one sample to the other. This is the kind of details that can decrease the “salability” of the products in the stores, or even cause a refusal from my client’s customer (the retailer). When I pointed it to the attention of the factory manager, he didn’t have anything to say. I expected a “cha bu duo” (it’s nearly OK), but this time there was no argument. Responsibility was totally diluted. Nobody really cared… “It’s not me, it’s the supplier.”

I often tell the factory employees they have too much work-in-process inventory. In general, they don’t understand how lower inventory can go hand-in-hand with faster production and lower costs.

Last week I was in a factory that really, really doesn’t see anything wrong with WIP inventory:

Another problem is that WIP can be lost, stolen, or damaged. This time I got a great illustration… I couldn’t help taking a picture:

They are sleeping on the fabric, before it’s used by the sewing operators. Isn’t it cute?