Blog Post

5 lessons I learned at Apple about how to design and build hardware

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Manufacturing beautiful and functional hardware is more difficult than ever due to capital demands and a lack of hardware experience on most startup teams. My experience at Apple taught me some important lessons about hardware design and production that, if heeded by hardware startups, provide an opportunity to bring innovative products to market without suffering setbacks – or even failure – from preventable mistakes.

1.  Get inside the factory

I’ve met too many people in this game who make one trip to China, pick a manufacturing partner, and never visit again. All of the companies they represent eventually end up with enormous problems when it comes time to launch. Get out on the manufacturing floor and talk to the line managers and operators. You’ll be amazed at what you learn about the manufacturing process and about your manufacturing partner. Seemingly small pieces of information from the factory floor can later help you refine product design for better manufacturability or even clue you in to larger issues with factory management.

In 2010, we had a supplier in China that had agreed to increase capacity by 50 percent over the next nine months. It had been a few months since anyone from my team had visited the factory, so I stopped by to see how the expansion was coming along. To my surprise, only about half of the new equipment needed was actually on order. After speaking with the floor manager, we learned that he was not given the resources to meet our expansion plans. Needless to say, we had to have a fairly direct conversation with senior management to get the expansion back on track.

2. Build prototypes close to home

3D printing houses and rapid prototyping shops are popping up all over the U.S.. You often get what you pay for in this realm, so it isn’t where you want to pinch pennies. Use the proto phase to refine, refine and refine some more. That way, when it comes time to spend money on pricey mass-production equipment, you only have to do it once. Mass production molds for plastic parts can cost upwards of $50,000, so finding out two parts just don’t fit together quite right after you’ve started mass production is an excellent way to jeopardize and even kill your company.

The added benefit of prototyping close to home is that your engineering team can iterate faster.  3D prototypes can be on your doorstep in a few days, compared to the four to six weeks needed for other prototyping methods. Expedited turnaround times accelerate overall development cycles, and in turn, reduce development costs. Rapid development also gets your product to market faster than the competition!

3. For mass production, China isn’t the only game in town

Examine the total cost of your supply chain. If you’re building product in China, you need to do the math on how it is getting to the U.S., where it will be packaged, cost of import duties, what happens if a product is defective, and a thousand other questions. Each of those factors has a cost implication, and when added together, startups sometimes discover that Chinese manufacturers are not price leaders after all.

Look for manufacturing opportunities closer to your customer. For instance, there is tremendous manufacturing talent and capacity in places like Guadalajara, Mexico, where you can benefit from NAFTA tariffs and reduced logistics costs, not to mention low cost of labor.

4.  The job doesn’t end after launch

Once you launch (congrats!), resist the temptation to sit back and watch it all happen. To the contrary, monitor your supply base like a hawk. There is a reason Apple has thousands of supply chain professionals on the ground in countries around the world. When things go wrong, they can go very wrong very fast. Actively monitoring supply chain data and maintaining a transparent relationship with managers at each node in the supply chain will prevent most issues.

When it comes to tracking data, inexperienced startups are often overwhelmed with the amount of data and tracking options a modern-day supply chain produces on a second-by-second basis. It doesn’t take an Apple-size team to avoid most supply chain issues. Figure out what your key data points are and track those on a daily or weekly basis. Take the time up front to build reporting tools that make it easy for you or your team to see at a glance if there is a problem building.

For instance, my team was able to monitor over a billion dollars of annual procurement across 22 factories using just six spreadsheets. Careful planning and foresight will go a long way towards ensuring that data can be used to proactively identify and resolve issues.

5.  Tim Cook is right – inventory is “fundamentally evil”

Most startups I’ve encountered are unaware of how excess inventory can quickly crush a small business. The simple answer to inventory management is to never carry more inventory than you absolutely, positively need (easier said than done). Before production starts, set realistic goals for inventory turns and days of inventory. If inventory exceeds pre-defined levels, shut down your supply chain. Shut it down entirely.

You simply can’t afford to have more product coming off the line if you’re not going to be able to sell it. You may find yourself in an uncomfortable position with your supply chain, but that discomfort is minor compared to the pain of writing off a massive inventory. If you don’t agree with this approach, please refer to RIM’s colossal $485 million inventory write-off at the end of 2011.

Oh, and one more thing:  Fire any engineer that ever says “it’s not possible.” That no-can-do mentality has no place at an innovative startup. Attitudes are infectious and that one is positively poisonous within an engineering organization that strives to innovate. People who are motivated by the challenge to push a manufacturing process to a smaller tolerance or a larger scale than ever achieved before are the lifeblood of innovative hardware organizations. Everyone else is just dead weight and a liability to your mission.

Bill Banta is currently a student at the Stanford Graduate School of Business and CEO of Stealth HD, which builds 360-degree video technology for the military and media broadcasters. Previously he worked at Square and also at Apple.

12 Responses to “5 lessons I learned at Apple about how to design and build hardware”

  1. Firing an engineer that tells you “it’s not possible”? Interesting… I agree that any negative attitude must be stomped out quickly, but that sounds like something from the MBA playbook.

    As a manager, I would think that you would prefer any engineer to tell you the truth, rather than what he thinks you want to hear. I would fire the ones who tell me only what they think I want to hear.

    With the truth you can adapt and evaluate your options in the real world and survive as a business. The alternative is to listen to what you want to hear for a while and get jerked into reality a few months down the line – by that time the engineer has realized the wheels are coming off and would have left by then leaving you with the mess.

    There is a difference between “it’s not possible” because it isn’t and “it’s not possible” because I’m too lazy, negative or just don’t care…. :o)

  2. Unfortunately, Design for manufacturing and design for serviceability are often at odds with each other. There is room for both, but especially in a market for ‘smallest/thinnest’ and ‘cheapest/most integrated’ there isn’t room for screws, or modular internals. In my industry, we have gone from very modular, very serviceable devices to single boards and multifunction custom IC’s. Makes service very difficult, but if you can get manufacturing cost low enough, the item becomes a ‘throwaway’ from maker’s viewpoint, even if the price is too high to throw on the consumer side. Very wasteful of materials, but in a bottom-line world it can produce the best margins.

  3. Wait, you say you design hardware for Apple? How’s about adding a #6 (or maybe even a #0): STOP GLUING YOUR PRODUCTS SHUT.

    Apple consistently ranks among the worst for serviceability of their products. There’s no excuse for this. If Apple wants to be green and eco-conscious, the first step is making your products straightforward to service so they can stay out of landfill in the first place.

  4. Simon Kwan

    Getting your engineers and designers to take on a challenge and have a ‘can do’ attitude is not all that hard, but getting your manufacturing partners to share your enthusiasm and vision is quite another thing. I’ve been involved in China manufacturing & sourcing for 7+ years now. I can’t tell you the number of times a Chinese engineer has uttered the words ‘bu ke nen’ or ‘mai ban fa’, which roughly translates to ‘it’s impossible’ and ‘there’s nothing that can be done about it’. It’s almost a standard Chinese reaction to anything they’re presented with that they aren’t already familiar with. It takes a fair amount of effort to illustrate & demonstrate that the things we want are, in fact, possible. When they finally get it, though, it’s great for both parties because I get what I want and they learned something new. That’s just the tip of the iceberg when learning how to deal with Chinese vendors. Don’t even get me started about the ‘cha bu duo’ and ‘da gai’ (‘close enough’ and ‘should be’) mentality!

    • billbanta

      Great points, Simon! It often takes several months (and a few small successes) to help engineering teams at Chinese OEMs believe substantial developments are possible within Apple’s rapid development cycles. Engineering managers are critical to ensuring their teams do not become overwhelmed with the scope of the task at hand. Spending time to build strong relationships with OEM engineering managers, and helping them understand that their success is aligned with yours, will go a long way towards establishing a positive attitude moving forward.

  5. Mike Sanders

    What a great article I was not aware of the size of the write off at RIM just that there was one. It also makes clear how important Tim cook’s role was in his previous job and I think as a shareholder I can sleep nights knowing that he is in charge.

  6. Thanks for the inputs from experienced supply chain work.
    Any comments on what kind of on-line DATA was taken and sent back to Apple for analysis of yield or rel issues later? Or did Apple have visibility into vendor engineering database? SPC charts? Test and measurement raw data?

    • billbanta

      I can’t get into too many details without creating a whole new post on this topic. The short answer is that Apple, like most OEM customers, has open access to yield and reliability data. Given the complexity of the products, engineering teams at both Apple and the supply base collaborate closely on design changes and manufacturing techniques. Supporting an open exchange of engineering data is critical to ensuring on-time delivery of the final product.

  7. billbanta

    Good point, Mike. I agree that when taken as an absolute, the last piece of advice is not practical.

    However, I’ve watched engineers (who do know a lot more than me) generate new processes or techniques to make the “impossible” become possible. The engineers who approach the challenge with the “it’s not possible” attitude are the ones that can derail an ambitious engineering team.

    Providing support and sense of mission to an engineering organization as they try to break through the boundaries of what is possible is key to maintaining good morale and a creative work environment in the face of daunting technical challenges.

  8. Yeah, I once asked an engineer to make us a perpetual motion machine. He said, “it’s not possible.” I fired him on the spot.

    The next engineer that I asked said sure. We are now two years into the R&D and they are telling me it looks good. Although, I often wonder why hits to Monster and CareerBuilder have been increasing on our network the last few months. Oh well, I am sure that once we make the perpetual motion machine that I will be filthy rich.

    Tongue in cheek aside. Advice to fire any engineer who tells you some idea that you have is impossible is pretty stupid advice when given as an absolute. Sometimes they know a hell of a lot more about it than you do. I do agree that bad attitudes can be infectious and need to be dealt with quickly. However, firing an engineer for telling you the truth is also a sure fire way to make sure that YOU are the source of bad attitudes.

    • Hamranhansenhansen

      The general idea here is that it is the job of engineers to make the impossible into the possible. Otherwise they are just copying what somebody else already did.

    • I think a better way for the author to have said it is “Fire any engineer who tells you some idea is impossible without any data backing it up, having already researched it”. While many of us enjoy trying to shout out what is and isn’t possible from the armchair, you need to take a shot before truly being able to say whether something is possible or not.