Why IT Is MoreBy CIOinsight | Posted 06-09-2005
Round II: Readers Have More to Say About Carr's "IT Revolution"
Nicholas Carrwho is notorious within IT circles for his 2002 article in the Harvard Business Review claiming that IT no longer mattersstirred up CIOInsight.com readers with his predictions about The Next IT Revolution.
We published some of the thoughtful letters to the editor commenting on Carr's analysis. While the overall tone was negative, most readers agreed with at least some of the points Carr made.
What could we do but continue to let them have their say?
Re: Carr's Outlook on IT
While I agree with some of Mr. Carr's ideas, such as data center consolidation, I disagree with the idea that utilization rate is a solid measure of value for most computing systems.
Most PCs, networks and servers are designed to have very low utilization rates. Five percent utilization for a PC is about right, 50-60 percent utilization for a network is appropriate, 35-55 percent utilization for a server is acceptable.
Think if it like this: A computer is like a car. Most people spend a small amount of time in their car, so it has something like 15 percent utilization on any given day. Now imagine only being able to use public transit to go places. Sure, in New York City that is fine, but not in many other urban or suburban areas.
Imagine highways having to achieve 80 percent utilization before they get expanded; rush hour would be next to impossible. 80 percent or more network bandwidth utilization creates that same kind of backup in traffic. Imagine only having one corporate bus that had to pick up everybody in the morning and drop them off at night!
That is what a 75 percent or more utilization can produce on a server.
If we start to take away the tools that improve our productivity in the guise of being more cost-effective, we are just being penny-wise and pound foolish.
That kind of thinking does lead to a Revolution; it leads to a loss of productivity.
Thomas Stiehm, Principal
Carr Is Correct, These Posters Are Wrong
IT now serves the careers of its members. 15 years ago, why was there no "Chief Telephone Officer" in the executive suite? Because the bloody phone mostly worked, and for what you getthe ability to talk to anyone anywhere any timewas incredibly inexpensive.
How is it that IT has leveraged years over years of failure to deliver a stable, safe, low-cost, reliable business tool to the desktop, by asking for and obtaining a senior management-level voiceCIO/CTO? One can only assume blackmail and/or a lot of slick-talking BSers.
I have worked in Corporate IT for several clients. The plan there is: Buy big, safe brand-name "solutions," deploy them inside tight PM/Change Management, and make sure your rear is covered for the inevitable business failures. Nearly gone are innovation, hacking, responsibility and customer satisfaction (when we do have that, it is usually obtained by reducing the definition of "customer" or "satisfaction").
I am working on a paper now, "The Rise of Stupid IT," in the spirit of David Isen's "The Rise of the Stupid Network."
My solution is that IT must be split into two streamsand this solution is the antithesis of what the industry is calling for. I recently attended an event here in my city (Calgary), the point of which was how governance is what it's all about now. A room full of lawyers and CIOs replacing what used to be a room full of hackers and technologists. Again, was such talk ever used for the telephone infrastructure in a corporation 15 or 25 years ago?
And is the whole PC/IT thing not really a more sophisticated telephone?
1) Infrastructurethe hallmarks are that it is tiny, and is expected to use legacy systems and keep them running for very long periods of time. New products are introduced to the company WHEN they are ready, not on any external schedule, and not on demand of any business needs. Innovators and hackers work here. You do not outsource these people or functions. (See barrage of recent press on the failure of outsourcing in general.) This type of IT reports to whoever is in charge of Facilities.
2) Consultingprovided as a resource to business departments that have pressing needs that could be covered by technology. These consultants are managed and paid for from departmental budgets entirely. Project Managers work here. Business application developers work here. They report to business people. Any questions about them and their work that senior management has are directed to the business manager they work for. Period.
You may outsource these people at will, they work on a project basis. Experts may be drawn from Infrastructure on loan. These projects do NOT change the operation of infrastructure. The lines of reporting are entirely business-related, based on the project in question. If business managers don't know enough lingo to manage these resources, they have to fund "translators."
Smart business managers will of course learn the lingo, or smart IT consultants will learn the business and be promoted there. Companies that get what I'm saying will survive in the emerging economy. David was right about AT&T.
Maybe I'm right about corporate IT?
Why IT Is More
Important than Electricity">
Re: Carr and IT
As Perry Cherpes said [in a previous letter to the editor]: "Electricity is the same for every companyinformation is not." What doesn't matter in IT in the future is all the things that are essentially the same across companies: servers, routers, any hardware platform that's been profitable for at least three years (think vehiclesyou may make rules about buying only Ford or Toyota for maintenance reasons, but you won't have a whole department focused on researching horsepower needs).
What does matter is what distinguishes your company from the others. The bureaucracies above mine (we are a small cog in USDA) wanted us to identify the common data items so we could share info easier.
Examining the data shows that the differences in our data exactly match the differences in our missions.
Our data (the database structure) AND the way users interface with that data (entry forms, reports, and sometimes the specific end-user hardware (think UPS)) are what matters. That's essentially your IT/business alignment.
If it's important for competitive reasons that your IT is always the best in the world, you will have a hardware department (just like the NASCAR groups do); otherwise you're wasting your money and can get along with Best Buy and maybe a consultant or single staff member up to date on industry news.
It's similar with data entry and reporting. If you have to accept strange inputs from various sources, you will need a development group. Same if you need to crunch humongous hunks of data faster than anyone else. If everything can be done through standard off-the-shelf stuff (smart cash-register/scanner/cardreader), then just buy it and focus on your specialty.
Barnes and Noble cannot distinguish themselves from Borders based on how they accept money at their stores, so don't focus on it unless you can think of some way cool thing that would make it attract more customers and be profitable.
What distinguishes them is not their inventory tracking system or their employee time and attendance software, any more than would the fact that their managers drive Fords instead of Toyotas (unless it was advertised company policy to "support America" or "support the environment").
A pizza delivery service can make a competitive statement by using Porsches (speed) or Peterbilt Tractors (volume), or they can compete in completely different areas such as freshness of ingredients or variety of menu.
Carr is correct in that right now, too many businesses have "departments of electricity" even though they aren't competing in any category that really needs that department. Those IT resources are unnecessary and will go away, turning into contract management departments before being absorbed completely into Finance. The other IT resources are immensely valuable and will not go away and still leave the company successful.
Mike Moxcey, Computer Specialist
USDA APHIS Wildlife Services
Fort Collins, CO
Dear Mr. Carr:
As a CFO I can hardly find fault with your desire to reduce inefficiencies. The problem lies in identifying those inefficiencies. Your contention that 70 percent of IT doesn't matter is somewhat glib and akin to throwing out the baby with the bathwater.
The first order of business is to keep the business operating. This requires an infrastructure that is reliable and, often, redundant. Consolidated data centers clearly save money, but also introduce single points of failure. The cost savings must be viewed as coming with added risk to business continuity.
The second order of business is to make sure the infrastructure can support reasonable growth and doesn't become an impediment to accepting and delivering new business. In my 35 years I have seen many short-sighted decisions based purely on cost that ended up costing more because equipment had to be replaced or upgraded less than halfway into its useful life. IT planning must be plugged into the overall business planning and forecasting function with a variable cost component based on growth projections.
The third order of business is to make it efficient, i.e. reduce costs and increase profits. I have seen how technologists fall in love with the technology. I have also seen how they can't seem to scale the technology to the size and needs of the business. They seem to believe there is only one right way to build an infrastructure.
That may be fine for very large and very complicated organizational structures, but usually results in large inefficiencies for small to midsize organizations. Most technologists just don't seem capable of recognizing inefficiencies that are staring them right in the face.
Based on the three items above, I have come to the conclusion that the IT function should always be run by a technology-savvy and strategy-savvy business person. That is the real key to realizing the next IT revolution.
Frank A. Van de Kerkhove
WT Chen & Company, Inc.