The construction industry suffers from the same problem that most ERP Java shops and aerospace contractors do - the only way they can stay in business using their business model is to use cost-plus contracting (software/buildings/military jets) and customize the hell out of their products to rack up the largest bill/longest hours/institutional knowledge (it's their incentives - hence most software projects fail/military jet fiascos/building overruns).
This could also be because some of these programs are actually hard to make (and some are) but it's also due to the fact that they have to keep reinventing the wheel and don't transfer institutional knowledge/expertise across projects or scale out their time.
They'll pursue this through "lowest" bid contracts, sales teams, obscure knowledge (APIs/systems/building codes), "personal networks" and other inefficient and economically wasteful mechanisms to maximise their take.
Product companies don't have this problem.
Prices are non-negotiable, things must meet exacting standards, things must be standardized, their is no institutional knowledge and there isn't any customization.
You get what you pay for - and that's it.
Java shops and aerospace contractors are starting to meet their reckoning with the rise of SAAS apps and vertically integrated companies like Tesla/SpaceX/Solar city. Who wants a customized rocket when I can get an off the shelf fully tested and consistent Falcon 9 for ~$60 million (even less than what China charges)?
Construction is the next industry.
Products > services for the simple reason that they are more cheaply produced, easily sold, iterated upon and created.
Comments
The construction industry suffers from the same problem that most ERP Java shops and aerospace contractors do - the only way they can stay in business using their business model is to use cost-plus contracting (software/buildings/military jets) and customize the hell out of their products to rack up the largest bill/longest hours/institutional knowledge (it's their incentives - hence most software projects fail/military jet fiascos/building overruns).
This could also be because some of these programs are actually hard to make (and some are) but it's also due to the fact that they have to keep reinventing the wheel and don't transfer institutional knowledge/expertise across projects or scale out their time.
They'll pursue this through "lowest" bid contracts, sales teams, obscure knowledge (APIs/systems/building codes), "personal networks" and other inefficient and economically wasteful mechanisms to maximise their take.
Product companies don't have this problem.
Prices are non-negotiable, things must meet exacting standards, things must be standardized, their is no institutional knowledge and there isn't any customization.
You get what you pay for - and that's it.
Java shops and aerospace contractors are starting to meet their reckoning with the rise of SAAS apps and vertically integrated companies like Tesla/SpaceX/Solar city. Who wants a customized rocket when I can get an off the shelf fully tested and consistent Falcon 9 for ~$60 million (even less than what China charges)?
Construction is the next industry.
Products > services for the simple reason that they are more cheaply produced, easily sold, iterated upon and created.