We have our economy model based on material goods: companies turn raw materials and resources to products that users individually buy and use. These goods get worn and eventually break, so the user comes back to the company for a new one.
IT artifacts are NOT of this kind. Apart from changeable technical boundaries, an idea, a digitized book, a software can be accessed and used at the same time by unlimited people at the same time. They don't wear off, don't break by the usage; if there is an error, it can be fixed and the change can again be globally deployed.
Shared. Fixable. Ever-lasting. These fundamental features make IT a science, or art, but totally useless for business. I can tell you the story, how important it was to practically RAPE IT with business concepts, to get money into it and so reach its current state. Repeat: it is very important that IT became a business. However: what I call a well-architected framework follows the fundamental rules of the IT: transparent (shared knowledge for any user), modular (fixable by anyone, not only the creator), and environment-independent: the tools, OS, language, running platform will surely change, the system should survive that with minimum or no side effects (ever-lasting).
These are VERY hard requirements, out of scope of any budget for any actual task, so you have no chance to think in long term on this scale: the ROI guaranteed to be negative! And if despite of this, you create an (IT-term) well-architected component, you SOLVED a task, and that is WRONG for your business! But the IT definition of "well-architected", you can go to sleep: anyone can use it, and if there is anything wrong with your solution, others can also fix it, no one is tied to you.
The only way for getting a financial ROI is to be better than your competitor, who in general, knows and wants exactly the same. How can you do that?
The key to business success is that you have real invention in your system, that changes or breaks a rule that your competitor follows. But you can't waste your resources on inventing it, because that is a long, trial-and-error story, and your competitors will come up with their products while your engineers keep scratching their heads. The only way is that you buy (worse) or steal (better) an innovation, build a marketable product on it, and you are the famous winner of the game. If you do it in large, you can be the next Gates, Jobs - if not, you can still have your holiday on Hawaii. You don't have to care for the innovators, who simply suck while you "polish their dusty rock". Or pay them some cash because you are a noble person, but that is bad for your ROI.
No personal offense in this. This is only the picture of the world where sharks became the dominant race, in NatGeo-style.
Do you want examples to these hard statements?
Let's take a look at the automotive industry. Giant factories were placed to China to build cars for the western market with cheaper workforce and less rigorous environment protection, so greater profit. Of course, China started to make its own companies, but those cars were... not good enough for us. However, today they could buy whole car companies, and their products suddenly became much better. What happened?
With the companies, they have bought not only well trained, but experienced engineers, and the research logs. _They_have_bought_the_errors!_ The guide that can shade the grey opinions, estimate not only the price tag and the financial ROI, but the (for me) real attributes, like possible results of crash tests after changing some variants.
Please look around in Detroit for yourself, and watch the marketing films about the Chinese advancement (but forget about the air pollution they pay for this).
Want another example? I was laughing loud when I read this, it is so telling...
I have read somewhere that Motorola was working on a modular phone architecture. WOW I said: an innovation against business plans! For dreamers: a phone manufacturer is interested in selling their product to the SAME user they sold the previous one - is it good for any of them that the users can replace parts?
Then I read that Motorola was owned by Google. The article also mentioned that therefore the only company that is not absolutely controlled by the number of sold items (surprise: is it not only me to whom this is obvious?)
Now I read that Google has SOLD Motorola to Lenovo, but... yes, you are right: they kept the patents, and yes, today Google has a modular phone project! And the article mentions that Google lost a lot of money on the Motorola story - I see they have a killer weapon against all companies who create mobile phones. And their project is not from Kickstarter, but made by the guys who have been in the heart of the business from the beginning, and contains experience from all the failures!
I wonder how many people remember that Android was not Google, AdSense was not Google... but today those guys work at Google - or in Hawaii. The exceptional big companies that played a noble role in the story are out of the IT playground today, like Xerox, or Sun Microsystems (try to understand the importance of making Java with all reference and tutorials free - as a university student I was a criminal, having a "stolen" compiler and documentations that I could not afford).
OK, this was really long and philosophical, although it answers practical questions and explains "magical stories" to me. My fundamental statement for the end: I accept your ROI-based approach, and its importance in today's world. I also accept that I must adapt to it. But I say that the fundamental architectural requirements of a truly, technically well-architected system is AGAINST ROI goals; and ROI appears only if the system contains something against ROI rules, that is not payed (stolen or bought for marbles).
