You can’t be excellent at software, hardware and services…
The “Ectors Principle” says that it is impossible to be excellent at hardware, software and services. You have to pick one and at most two but you can’t have all three.
Why can’t you be excellent at all?
Hardware is a discipline that requires you to define everything upfront and then to produce in as large volumes as possible to get to the lowest price per unit. It is a volume business with low margins.
Services requires people that understand customers ‘ problems and try to find solutions for them. It focuses on integration, customization, support, etc. To double revenue, you need to double the workforce.
Software is the art of combining easy to use solutions, with extensible and customizable features at an optimized and acceptable level of performance and availability. Smarter programmers can win from armies of dumb programmers.
So how can you combine a hardware platform that needs to have all features prebaked, with a people business that focuses on adding new features for each customer, with a software in which the art is to offer more with less features because feature overload kills ease of use, performance and availability? The simple answer is you can’t do them all well. You have to pick.
Who is trying to prove the “Ectors Principle” wrong?
IBM, HP, etc. come to mind when the words hardware, software and services are mentioned. How are they doing? BAD!!! Please name 5 HP software products that you find inspiring! IBM at least realized that being excellent at all three is hard and sold its PC and server business to focus on services and software. A word of advise: spin-off Power processors as an independent company!
Apple? Apple is not in the services business because you can’t have a closed software and hardware ecosystem and try to customize it to everybody’s needs.
Google? Google tried hardware with Motorola (failure!) and Nest but is not focusing on services because they don’t want to have to hire double the amount of people to double revenue.
Amazon? Hardware is proving difficult. Enterprises cloud services was a stretch.
General Motors? Do you like the software that comes with cars. Would you call it excellent?
Cisco, Ericsson, Huawei, Nokia, etc.? Software defined networks will kill their hardware margins in the next months/years. Alacatel Lucent just threw in the towel. More consolidation to follow.
General Electric? Wait until everything becomes “smart”. If GE is smart then their smart fridges have app stores with apps and peripherals from others. If not it will not be pretty.
Focus on one
Ideally you focus on one discipline and have another one as a hobby at most. Trying to do all will result in a weak link in the chain. The exact link that competitive startups will focus on to bring the imperium on its knees…