7 ms·
From a practical perspective, it seems to me that the Beer Game rests in three parts: in imperfect information, time delays, and most importantly, independence
by kevinalexbrown 15y ago
From a practical perspective, it seems to me that the Beer Game rests in three parts: in imperfect information, time delays, and most importantly, independence of agents. If each member of the supply chain knew consumer demand perfectly far enough in the future, they could compensate. Likewise, if each member of the supply chain could instantly scale their process, they could compensate. The key, though, is that even if you perfectly forecast, or can instantaneously process and ship your stock, everyone down the line from you must do so as well.
From what I gather, having the supply chain in China or Brazil addresses the second two problems. The advantage of China isn't just cheap labor, it's cheap, readily available labor. A change in demand (like switching the screens on the iPhone) can be met relatively easily, which partially solves the time delay problem. And the more of the supply chain you hold in one geographic location makes shipping time faster. The independence of agents issue gets addressed when Apple can coordinate the actions of each member of the supply chain better. I imagine this is, for the moment, easier in China than the US.
Where I part ways with the article is the bail-out. The US labor pool doesn't seem elastic enough to address the time-delay part of the bullwhip effect. If Foxconn suddenly had a temporary drop in demand and had to lay off thousands of laborers, it would not be a huge issue. If they need to hire them again, zip, pretty quick. Likewise if they need to request huge overtime commitments because they can't train new workers fast enough, no problem. US workers are particularly averse to uncertain job prospects, perhaps with good reason, but with the effect that scaling work (and wages) up and down is less tenable. And perhaps this is where the unions come in - the overtime payments from forcing workers to work more hours to meet demand is much more costly in a union than a non-union environment, for better or worse.
Edit: I've just now recalled that one summer, I worked for a large bread factory. They loved hiring college students in the summer because they knew that they didn't really care about staying in the union. Summer demand for baked goods skyrockets. A plant manager showed me how they were lagging behind demand by something like 2 million dollars / week (I've forgotten the exact number, but it was staggering to me). In his words "you can imagine what that does for business." But they couldn't just start hiring more workers, or scaling up the plant, because then they'd have more workers than the knew what to do with once the winter hit, when people were working 20-30 hrs/week. And this is one case where almost the entire supply process was in the US, minus wheat they might have gotten elsewhere.
- excuse-me 15y agoYes if you knew the number of iPad 5s that people in each city in the world were going to buy in 2years time you could scale the production in advance and it wouldn't matter how many delays there were in the supply chain. Since you don't know the demand - you need a responsive, high bandwidth, system. Having suppliers that can't respond quickly - either because they are a 2week container ship voyage away, or they take 3months to respond to quote requests, or they think they can make more money by demanding huge penalty clauses for any changes in order - all these add resistance and slow the response. There is also a cost consideration. Once you start making a component it is costing you money until the consumer pays for it. Having to make a billion $ worth of screens/chips/cases 2 years in advance is costing that whole network of suppliers money - costs which will be passed onto the customer
- Retric 15y agoThat's one piece. The other huge issue is that a Chinese worker can produce an iPad for less than a robot while an American worker costs more. A lean supply chain looks great on the books, but it's also a huge bet that nothing bad happens. However, add flexible enough automation like say having 500,000 humans and you are far less dependent on any one customer so you can ramp up and down vary quickly. Which both reduces the need for a large supply chain and cuts the response time to changing market conditions.
- scott_s 15y agoThe other huge issue is that a Chinese worker can produce an iPad for less than a robot while an American worker costs more. Not by as much of a margin as one would think. Listen to part 3 of This American Life's retraction of the Daisey story: http://www.thisamericanlife.org/radio-archives/episode/460/retraction http://www.thisamericanlife.org/radio-archives/episode/460/r... The conclusion is that the difference per iPad or iPhone is a small percentage of the profit. They argue that it is, in fact, the supply chain that makes the difference.
- Retric 15y agoThe estimate was moving manufacturing in the US would cost up to 65$ per iPhone and Apple is selling around 80 million iPhones a year. That approximation 5 billion dollars in lost profit each and every year just to build iPhones in the US. That's simply untenable for a publicly traded company even ignoring the added cost for iPod's and iPad's. Now they could probably give up say 100 million in profit per year and have dramatically better working conditions in China, but even that's a hard sell to stockholders. Having said that, Apple could redesign the iPhone to be easier for robot's to assemble and move manufacturing back to the US, but it's simply untenable to have Americans assemble such things at this scale in the US. But, this is where the supply chain issues show up and if they designed the iPhone to be built by robots they would as you suggest probably locate those robots in China.