Wednesday, October 27, 2010

Data volume and variety

At a meeting with one of our main clients’ referral the topic of data security shifted to the amounts of data at their disposal, therefore being very agile when it comes to obtaining reports, statistics and such. One of the main decision makers was pointing to the fact that they had close to 100 GB of data “perfectly safe” thanks to the systems and technologies acquired as of recent.

At one point during the discussion I asked what tools they were using for validating the wholeness of such data, making sure there were no duplicates, which interfaces they used to match diverse types of data and their respective packages to talk to each other, how they obtained updated statistics from such data, and a few other questions.
I started projecting the following:

“What if your data were not as c mpl te as y u t ink it is?”
“How do you make sure some of it do you make sure some of it is not duplicates is not duplicates?”
And so on…

Turns out the diverse types of data they posses are not standardized and most programs used have no way to communicate to each other. So, some information stored in database A is duplicated in database B; while some of such data is duplicated inside the same database. There is no policy or tool that guarantees the data is retrievable from backup and even from its original repository. And there is not Business Intelligence package or procedure to gather information from all these places to have instant reports on the financial health of the business or any other operation or department.

In other words, this is another instance of what we refer to as data rich, insight poor.

It was obvious that the IRR on all these expenses (not investments in this case) was of a negative value, an infinite amount, but in terms of time.

The main issue?
Business-type decision makers pointing to the IT-type ones as “not getting it.”
IT-type decision makers pointing to the Business-type ones as “not learning it.”

What is the catalyst for Marrying IT with Biz?
Training.

The least we can do is sit down the biz people for a day listening to IT: not about the technology insights or anything in terms of Terabytes, Gigahertz or Operating Systems; but rather on the reasons why a particular technology is better than the “inexpensive” or “popular” or “affordable” ones; and how those integrate –or not- with the rest of the existing applications.

And then sit down the IT people for a day listening to Biz: not about the budget projections or anything in terms of Marketing, Cost Containment or JIT Inventory; but rather on the reasons why a particular process is better than the “easy” or “accepted” or “cool” ones; and how those incorporate –or not- with the rest of the existing departments.

Once IT gets it and Biz learns it, all that data in your possession will be better able to provide information; its sole reason to exist.

Ready for the meeting?