Data is money

NEW ORLEANS -- Make big profits now with data warehousing. Yowsa, yowsa, yowsa. ROI will be oh, so high. Well, that wasn't quite the thrust of the NEXPO '97 workshop on data warehousing, held Monday, June 23, but that's about as exciting as the presentation can be made.

Data warehousing entails more than the electronic equivalent of carting paper records into a big building with lots of shelves.

No, in these days when information is power, and information organized in a database is even more power, and the ability to reuse (repurpose, recycle) that information is the equivalent of grabbing the brass ring. Data warehousing is important, particularly for organizations that have oodles of very disparate data, like large newsgathering organizations.

Even more important for organizations' health is the Year 2000 problem, which was also outlined at the end of the workshop by David Lightfoot, president of Annapolis, Md.-based Chesapeake Research Group.

"It's not like the national debt, you can't pass it down to the next generation," he said, noting that the problem is more of a concern to newspaper accounting and business software than it is to editorial systems.

Newer computers running Windows 95 and NT, and all Macintoshes, from the oldest to newest, aren't affected by this problem, but it still leaves a sizable number of accounting and financial systems, and legacy systems, in need of revision or fix.

After a brief introduction from moderator Elizabeth Sholar, director of publishing systems at Thomson Newspapers of Stamford, Conn., the first speaker proceeded to define the scope of data warehousing, in a jargon-laden presentation that benefited from the handout defining terms at the end.

John Fee, practice manager at Thomson Technology Consulting Group of New York, enumerated four types of data warehouses, which contain multiple small databases to large all-inclusive databases. To make these databases useful, you need query tools. Then you can go mining for data, using algorithmic applications with the ability to rank search results and other "machine learning" functions, as he referred to them.

It all boils down to: data is money, companies can get a 40 percent return on investments in data warehousing and the payback time on investment is 2.3 years.

With a real-world example came Lorie Schrader, the applications development manager of the Dallas Morning News, who described how the paper has leveraged their store of information in its data warehouse for targeted marketing of both subscribers and non subscribers in a given area with a free shopper and specialized inserts for that shopper. There were many benefits from this targeted approach, but it was found that the recipients all love to get junk mail about the junk they like. It's like "panning for gold," she enthused in this more focused, less jargon-y presentation.

Next came Don Stroud, management information systems director at the Plain Dealer Publishing Co. in Cleveland. His topic was data warehousing in a circulation and advertising environment. He favored databases that could be easily manipulated by end users, and not requiring "programming types" because the database use was too complex. There were problems when bringing together data from circulation and advertising into a single data warehouse, since both are so oriented to a production environment, but progress had been made.

Steven Hannah, the vice president, information technology for the Gazette Co., Cedar Rapids, Iowa, structured his talk around a case study in data warehousing -- his company, which has a multitude of data types, from TV, radio, new media, database marketing and a shopper, in addition to the daily newspaper.

"Our mission is to be the information provider of choice with a dynamic mix of innovative products and services," he said, and it seems as if the company is for 22 counties in Iowa, southern Wisconsin and Illinois.

Sounding a bit like a dentist, Hannah stressed the importance of data hygiene -- you could have "an electronic trash heap" without the proper care in placing disparate data in the warehouse. He also mentioned the usual security issues to prevent the data from being compromised.

The question of a Big Brother presence and individual privacy was mentioned, but Hannah stated that all legal and ethical questions were being respected, and that all the data gathered came from strictly public sources. Still, all Gazette employees are required to have a confidentiality statement on file before accessing the data. And a final, reassuring point: The Gazette will never sell the list of names it has warehoused, but will gladly take a client's money to use the list to do a mailing.

Leaping back to the Year 2000 problem, the best solution Lightfoot offered was to act today, while there was still a bit of lead time. Since many contracts that go beyond the Year 2000 will have to be written soon, the problem can't wait until 2000 to be solved. Adding to the complexity is the fact that 2000 is a leap year, a fact not mentioned in many discussions, but complicating the problem of synchronizing calendars previously based on two data bits.

On balance, the entire workshop was worth more than that, and was well attended, but nearly half in the hall had left before the Year 2000 segment finished. Perhaps what was last should have been placed first, since we all learned "time is money" before we knew that "data is money."

-- George Powell

A production of The Cole Group, Copyright (c) 1997, All Rights Reserved.

Last updated: 26 June 1997.

Return to The Cole Pages