From Data to Information

Greetings from Las Vegas. It is popularly said of this glitzy city that "what happens here, stays here." This post is an exception to that rule. IBM is communicating an important idea to about 6,500 people (myself included) attending the Company's Information On Demand 2007 Conference, and I want to communicate it further through my blog. The message is deceptively simple: in the world of data, what really matters is information. Data-oriented IT professionals - be they DBAs, systems programmers, or application developers - focus a lot of attention on database management systems such as IBM's DB2. These database "engines" are obviously important, but they are pieces of a larger puzzle, and the big picture is all about information.

So, what's the distinction between data and information? As I see it, information is data in a useful form. I'll illustrate the data-information difference with an example culled from my baby-boomer memory bank: quite a few years ago, when the "Saturday Night Live" TV show was just getting started, Chevy Chase was hilarious as the anchor of the fictitious "Weekend Update" news program. Once, in that role, Chase said that he had some football scores to report: "12 to 7, 28 to 14, 35 to 10, and 17 to 3." 12 to 7 is data. Dallas 12, Philadelphia 7 is information.

Now, you can't stop (at least, I can't) with the notion that information is data in a useful form. Information confers value on data, but the value of information itself depends on a number of factors, and people who work with database technology - in whatever capacity - would do well to keep these very much in mind:

• Timeliness - Take a look at the "front page" of the CNN.com Web site. You'll see next to some of the headline items an indication (emphasized in red) of how much time has elapsed since the associated stories were posted or most recently updated (e.g., 38 min, 5 min). The folks at CNN are no dummies. They know that information can quickly go stale, and they want their site visitors to see that the goods on display are fresh. The same is often true in a business setting: getting information quickly to people who need it can have a significant impact on an organization's performance. I like the way Arvind Krishna, IBM's Vice President for Data Servers and Information Management Development, puts it when he speaks of the "increasing velocity of information" - the need to get information ever more quickly from point of origin to point(s) of consumption.
• Accuracy - This is much more than simply guarding against erroneous database updates and inserts (although that is important). Informational accuracy is also about data rationalization. If you're a clothing company and you don't know that the Pat Richards in your customer database is the same person as the Patrick Richards who's also in the database, you won't be building goodwill when you send Pat a catalog of ladies' sportswear.
• Reliability - A company loses credibility with securities analysts if its earnings forecasts consistently turn out to be well off the mark. Is there a problem with the data on which the company bases its forecasts, or with the tools and processes that generate the estimates?
• Consistency - You've heard the words in countless commercials: "Has this ever happened to you?" People from an organization call five individuals employed by one of the company's suppliers, and get five different answers to the same question because the five people on the supplier end are using five different (and uncoordinated) informational systems. This kind of situation underscores the importance of having a "single version of the truth."
• Accessibility - Of course you want end users to have ready access to the information they need to do their jobs, but what about application developers? Can they access data in your system in a way that makes boosts their programming productivity? If they want to use JDBC or ADO.NET or Ruby on Rails, are they told, "COBOL or nothing?" If they want to use dynamic SQL to speed the delivery of a new application function, are they stonewalled? Is that right?
• Relevancy - Information about last month's electric bill for a store can be useful, but not to the person responsible for keeping the right product mix on the store's shelves. The right information has to get to the right people.

Sometimes IT infrastructure people draw the line at cost-efficiency, throughput, scalability, and availability, and figure that other imperatives (such as those listed above) are someone else's problem. The goals towards which these people devote their efforts are indeed important, but they are only part of story when it comes to providing the informational capabilities an organization needs for breakthrough success. Techies ought to realize that they can make valuable contributions within the broader domain of information management, and that they are on the same team as their colleagues who work further "downstream" along the course by which data becomes what the business really needs: information.