Jokes about Milkmen have been around forever, even if the practice of home delivery has been dying off.
I don't ever think my family used a milk delivery service, and as the picture here will demonstrate, I look just like my father thank you very much. After a peak in the early 1960s, the home milk delivery business plummeted until about 2000, and then started to pick up again. Mostly, it is estimated, that the combination of convenience for two-income families and the desire for fresh local milk has spawned a re-emergence of this industry.
There are only a few key elements to milk delivery business. Get them right - you stay alive. Get them wrong, hello Monster.com. Customers of home milk delivery want convenience and reliability. Frankly, they can buy milk just as good from the local grocer - and for less money. Everything you do, must be focused on convenience and reliability. Period. As Dad might say, nothing else matters.
If you were an Enterprise Architect in a milk delivery company you'd design systems to maintain the tenuous just-in-time stream of fresh milk from local suppliers, schedule drivers, configure delivery routes, drive down costs, maintain infrastructure, lease vehicles, manage inspections and repairs, and of course, conform to regulations.
Imagine pitching a new GPS tracking system to improve on-time delivery and enable real-time tracking of milkmen. For the first time ever you would be able to monitor milkmen who stay suspiciously long at certain addresses! Snicker, snicker.
If you think you want the new cool GPS, you have to equate it to convenience and reliability. I don't mean you have to find a way to cleverly convince your boss that a GPS will improve reliability - you have to be able to prove it to yourself; and you should be a skeptic.
As technologists, we can be overly enamored with the prospect of adding new solutions, tools, and capabilities to our arsenal.
Architects must understand the key drivers of our businesses and skeptically evaluate every solution in terms of it's impact on those key drivers. Are you concerned that your company doesn't have the latest technology? If you can't prove a need, then spend your intellectual energies on improving your use of existing technologies.
Here's a test -- whatever the cost of your next project, assume it is the last 'N' dollars you have for the next two years. What project would you spend it on? Before you implement that SOA solution, the latest JDK, the newest database engine, or a cool new network appliance, prove there is no better way to invest the money.