Does IT Matter? 363
geoff313 asks: "I'm sure many of you are aware of the uproar over Nicholas Carr's article 'IT
Doesn't Matter' which was published in the Harvard Business Review, back in May. While many big names in the IT world have responded already to Carr's article (Ballmer has declared it 'Hogwash'
and Fiorina has pronounced it 'Dead Wrong'),
Carr debated vendor executives Monday at the Comdex trade show, proving that the issues he raised are still resonaating through the industry. Do you feel that corporate IT budgets
should be focusing on cutting edge technology to best serve its customer's needs, or should they focus on shoring up what they have now in order to maximize its usefulness to the customer? Some background can be found from the Washington Post,
InfoWorld,
and ZDNet, as well as at Nicholas Carr's site."
"For those of you unfamiliar his philosophy, it can be summed up pretty thoroughly by his statement 'Follow, don't lead,' arguing that the huge advances in the IT industry over the last two decades have erased the strategic advantage to be had by corporations for staying at the cutting edge of technology. In short, he advises 'executives need to shift their attention from IT opportunities to IT risks - from offense
to defense.' Of course the head honchos at IBM and Microsoft disagreed with him, citing Wal-Mart's use of RFID tags to keep track of inventory and other forward thinking IT decisions as a refutation of his thesis.
What I am interested in is the opinion of those in trenches of the IT war."
The answer is both, duh. (Score:5, Informative)
When the question is whether to boldly lead, or cleverly follow, the answer is always both. You lead where you can, where you have opportunities to, because your IT department taking some initiative in expansion means that you can grow the business above it. You have resources, products, and customers, and IT sits in between all three of them to some degree, and makes them possible, just as your maintenance department does. After all it's kind of hard to have meetings if the lights are off, right? And it's hard to do business when you can't get to your databases, or if your customers don't know about your products, or whatever else that isn't possible without IT.
The solution is always to strike a proper balance between expansion and consolidation in all of your departments, lest they grow too large and consume too much of your resources, or fail to grow enough to keep up with the rest of the company. It doesn't matter if we're talking about IT or R&D.
True but.... (Score:5, Informative)
Basically, companies don't want to change the way their fundamental "business processes" work even when these "processes" don't make any sense. So if you take the same old inefficient way of doing things, and make software to facilitate it, you're still doing it inefficiently. Especially when requirements for "visionary" systems get bogged down with specification by committee - everybody wants to make sure that their department or group level jobs are represented and that nobody designs them out of the picture. Even if a top level executive recognizes that the way things works is too costly and generally sucks, if lots of mid-tier shitheads play the bureaucracy card and bog a system down until it's in le toilette, well, no surprise when the software you end up with is no better than the way you do things now.
It also doesn't help that "IT" is the result of years and years of evolution and almost NOBODY in the business IT world is sufficiently bright to take the big picture, generalize about it, and create a logical, functioning infrastructure to replace it. No, the people who are smart enough to do this generally work for tech-focused companies in more interesting jobs where there are tiers upon tiers of bureaucratic wretchedness breaking everything down.
Re:Balmer (Score:4, Informative)
Case Study (Score:3, Informative)
Three years ago we had a Solaris network, Sun workstations, Netscape mail and calendar server, etc. There were five sysadmins for 1200 employees. An "Intro to UNIX" class was held twice a year, and an "Advanced UNIX" class once a year. All in all, it was a traditional, stable, robust, and boring infrastructure.
Then we got bought out by a huge multinational. We went Microsoft-only. We're now on a Windows network, Win2K and WinXP workstations, Exchange server, etc. We have 20 MCSE's for 1000 employees. Introduction to Windows classes are held quarterly, with additional classes in Word, Access, PowerPoint, etc. Our network is now unstable, frequently down, and very exciting. We have to reboot our workstations to apply patches about twice a week.
If it ain't broke, don't fix it!
Re:Easy test (Score:2, Informative)
If you stop any large established system straight away, of course there will be chaos.
Take the sewer system for example, if you filled up all the pipes overnight, there would be a horrible mess and chaos. That doesn't mean we could not function without a sewer. If we returned the poo vans along side the garbage trucks, embarked on a major education campaign and did a lot of house renovations, there would be grumbling and complaints, but the old way would be gone and the new (or even older) way would be working and nothing would have to hit the fan.
Replacing any system, even over time, is a very hard task, but it can be done. A business can run without computers, but it probably wont run as efficiently or as profitably as it did with them.
The way forward... (Score:3, Informative)
X Terminals were a great idea but in a time where machines and network infrastructure were too slow to support them. They have pretty much gone away.
Today your average desktop class machine is really enough to support several dozen regular business users.
Add openMOSIX [sourceforge.net] into the mix, and one virtual machine made up of a small handful of real machines can suddenly support hundreds of users' desktops. New machines can easily be rotated into the cluster (live) while old machines are rotated out when they become obsolete.
On the actual desk itself, something like a VNC terminal appliance is all one needs. Lifespan of one of these units is several times what a PC would last.
A sysadmin with 300 users is now really supporting only one workstation (whose processes are being migrated to maybe a dozen or two other workstations who have direct access to the master node's file system).
This isn't pie in the sky. It's based on very old ideas re-applied using new technologies that weren't available when the ideas were first tried. It actually works very well using the hardware and software available to us today.
I have to laugh when my users think that what I'm doing is bleeding edge. This is old school UNIX administration.
It's the ECONOMY stupid! (Score:3, Informative)
While most people thought it was magic unleashed on the world, I saw the Internet as a developed captive military technology released into the commercial sector. It was like a grant from the DoD. Is there more where that came from?
If you can sit back and receive defense technology and all you have to do is product development, then of course it makes no sense to burn cash on long-term R&D. Many economists and political scientists think that R&D is more efficient and effective in the public sector anyway. R&D in the private sector yields patents, which end up limiting the economic impact of the technology developed.
I can swallow Carr's ideas if he means companies should pool their resources into a creative commons for serious R&D, and then everyone can share the intellectual harvest. I can't swallow his recommendations if he means that advancing IT is too risky or otherwise unprofitable for anyone to bother with.
Ironically, IT has no value when you don't know how it reduces waste and generates economic benefits. If you do what Carr recommends, sooner or later you will think yourself right through the "IT is pointless" argument because you will understand the needs and know your idea coming from practical business operational knowledge, is a good IT risk.
I'm not saying I agree with Carr, but I wholeheartedly agree with him promoting this kind of debate! Also, are the people in commercial sector IT qualified to distinguish between a good idea and a bad one? I still believe we have a serious shortage of smart people. I think we should slow down IT to the rate at which smart people can deliver good IT.
another opinion.... (Score:2, Informative)