Learn to Manage Data, Not Crises

Yes, it was probably insensitive of me to tell a friend who was recovering from heart surgery that operating on a major organ is not only inconvenient and painful, but it's also the crudest, most temporary, most incomplete treatment for any systemic problem.

Hey, I was trying to encourage him! My point was that there is always (eventually) a more elegant alternative to radical intervention, even if it's not available to you at the moment you need it. That goes not just for medicine, but for IT as well.

In medicine, the first response to any serious problem is almost always surgery, at least until less-violent ways to deal with a problem are discovered. In my friend's case, that meant inserting a catheter with a balloon that was inflated to crush plaque against the walls of an artery. That's relatively low impact, considering that the alternative would have involved cracking open his chest, but it's still kind of a Roto-Rooter approach.

In IT, you get the same crude responses to problems like the overwhelming buildup of corporate data. The two basic approaches are to just spend endlessly on storage or to lop off huge chunks of excess data and then impose random disk limits that might keep people from doing their jobs.

There are some in-between approaches, like making storage more efficient with storage-area networks, or layering XML or other middleware on back-end databases so you only have to maintain data, not scrub and then warehouse every gigabyte individually. But those just put off dealing with an inevitable crisis—a process at which IT is quite skilled, but which is much less effective than preventing the crisis in the first place. Luckily, there are specialists who have the skills to deal with deadly data buildup—if you know to call on them.

They're records and information managers, who are only now emerging from the dank warrens of accounting and legal departments in which they developed their arcane skills and ruthless attitudes toward records that have outlived their usefulness.

IT has never had any real discipline about how electronic data, spreadsheets and Word files are stored; IT pros have never been able to act in an informed way about the content or value of the data, because IT deals only with files, not information, according to Angela Fares, corporate records and information manager at Radio Shack Corp. in Fort Worth, Texas.

Effectively managing information means knowing when it was created and why, what its value may be and to whom, in what medium and for how long it should be maintained, and when and why it should be destroyed, according to both Fares and ARMA International, a global information and records management association. (Its Web site, www.arma.org, has more information and guidelines.)

That means conducting information audits, pushing departments to define what information they want, what they want to do with it and when it will no longer be of use, Fares says. It also means that IT, legal, accounting and records management people need to collaborate on a cohesive policy for defining and managing data, not just deciding how much money to spend on disk space.

That's a significant change from the way most organizations handle—or even think about—the information they collect. But it's one that can keep them from being strangled by their own data. It can also keep them nimble and healthy by making sure that they're making the right information, not just every bit of data, available when it's most needed.

It's the elegant solution to data glut, the healthy approach, the choice of the tofu hot dog over the rare cheeseburger—solving a problem by making sure it never becomes a crisis.

But (and you'll have to trust me on this one) it's not a solution you want to suggest to people who are still recovering from the crisis in the first place. Somehow, just when they're backing away from the brink of disaster, some people are just not interested in how much more gracefully you would have handled it.

Kevin Fogarty is a freelance writer in Sudbury, Mass. Contact him at kevinjfogarty@yahoo.com.



Benefits Sought

IT managers at large organizations identified the following as the most important benefits they’re seeking from data warehouse projects:


(scale of 1 to 5)

Business performance mgmt. 4.0
Financial analysis 3.9
Customer analysis 3.7
Production/operations analysis 3.6
Workforce analysis 3.2

Base: Survey of 264 IT managers at North American companies with 1,000 or more employees

Source: IDC, Framingham, Mass., December 2001

Business Headaches

The top five business challenges for data warehouse projects in large organizations:

1 Cost
2 Executive support and company buy-in
3 Legal/regulatory and industry standards
4 Logistics
5 Staffing issues

Base: Survey of 264 IT managers at North American companies with 1,000 or more employees

Source: IDC, Framingham, Mass., December 2001

Data Challenges

The four most difficult types of data to deal with in data warehousing projects:


Legacy data 32%
ERP data 18%
Internal data (e.g., spreadsheets) 18%
Internet data 10%

Base: Survey of 112 IT managers

Source: Cutter Consortium,Arlington, Mass., December 2001

Special Report


Taming Data Chaos

Stories in this report:

Copyright © 2002 IDG Communications, Inc.

It’s time to break the ChatGPT habit
Shop Tech Products at Amazon