And of course, established hardware players know that their advantage in components is a huge barrier to entry. Some of them, such as Sony and Samsung, are willing to sell the same components used in their products to other companies as long as they do not compete directly with them. Other companies, like Canon, make their own components to stay ahead of the competition. Meanwhile, Apple directly buys companies that produce its key components.
Another reason these established hardware companies have been reluctant to resell their components is because they market their “specs” or technical specifications. Impressive hardware specs attract media attention and reach millions of customers who don’t really understand the performance differences. This behavior is only reinforced in the tech media narrative when companies like Apple market things like their A7 chip or Nokia tries to come back with “41 megapixels” in its smartphone camera.
>Lack of automation forces small startups to become global entities before even reaching the market.
What if new software companies could only compete if they created new types of code running on old servers and so Did you have to market your products based on the specifications of your servers and the quality of the code itself? That would be the end of the lean startup revolution. But that’s essentially what’s happening with the hardware and components.
The only way the competitive landscape around components will change is if new companies break the monopoly or if big players like Apple, Samsung and Sony become more willing to sell their components. This last path limits the path to success for hardware startups, so expect more startups like Electric Imp (which produces a wi-fi chip with cloud infrastructure) and Ambarella (which produces a video processor) obtain successful financing to manufacture components for everyone.
Software is already eating part of the world. But for hardware to be successful, software also needs to consume manufacturing.
Marc Andreessen was right: software is eating the world. But if it doesn’t also eat into manufacturing, hardware will forever remain a money-consuming, volume-demanding machine.
The well-documented delays of almost all crowdfunded products tell us that manufacturing is difficult. Incredibly difficult. If hardware startups have any hope of competing with the big, established hardware players, we’ll need a new generation of vendors: one that is automated and real-time.
Currently, suppliers’ reliance on manual labor means they can only work on products that leverage the same labor, processes, and testing equipment. This makes providers susceptible to becoming obsolete when their category of expertise becomes irrelevant or when their customers grow beyond their capabilities. The lack of automation also limits hardware production to countries with cheap labor, forcing small startups to become global entities before they even reach the market.
>If software doesn’t also eat into manufacturing, hardware will forever remain a money-consuming, volume-demanding machine.
Instead of relying on six-month-ahead forecasts based on non-existent data, these suppliers also need to be able to make the 500 items a hardware company sold yesterday and have them delivered to the customer’s doorstep tomorrow. This is a huge challenge, which also requires all other moving parts, including the entire component supply chain, to be delivered in sync.
Finally, suppliers need new tools that allow them to work more closely with their customers. Take a software analogy: The current state of hardware designed in the United States but manufactured abroad is like separating front-end developers from back-end developers by thousands of miles and double-digit time zones. That may not seem like a big deal in software, but in hardware, the chasm between design and manufacturing can result in defects costing millions of dollars and endless delays.
