* Data classification projects don't need to be painful
Why IT projects succeed or fail sometimes has no direct relationship to the quality of the project itself. Sometimes projects fail because nobody properly attended to the requisite politicking. Consider, for example, a very typical assignment in IT these days, a data classification effort.
By some accounts, up to half the data on corporate servers is aged, or for some other reason is of little or no value. Some companies report that much of the data on corporate storage devices is actually private data, and has nothing whatsoever to do with the company’s business. Why then should a company be expected to provide storage assets and storage services to protect such data?
In most cases, it should not.
And so occasionally, an IT manager chooses in the course of implementing one project or another – perhaps an ILM project, but it could be anything else – to spearhead an effort to understand the value of data. Understand the value of your data, and the assumption is that by moving lower value data to cheaper assets, you can ensure there will be plenty of room on the expensive arrays for the high-value data.
The road to hell is paved with good intentions.
In the 1950s, Deming tried to make corporate America aware of the need to get buy-in from everybody involved in a project. “Everyone” of course meant … everyone, a group Deming referred to as “stakeholders.” In the case of the data valuation project, stakeholders would be anyone with data on the corporate storage systems, the IT group that manages those systems, and so forth. Everyone who would be touched by the project should have some appropriate input.
You may recall that in the 1950s, America didn’t listen terribly well to this advice, but Japan did. After several years, even though in the U.S. we may not have been listening to Deming, we sure noticed what the Japanese were accomplishing. And that, we paid attention to.
Today, we all understand the importance of stakeholder involvement, but some project implementers still just go through the motions. And that’s a pity, because these meetings with the stakeholder groups offer opportunities far beyond data collection. First, they are a chance to put a human face (see how I give my readers the benefit of the doubt!) on IT. Second, they are a great way to get managerial buy-in in front of the troops.
Get the managers on your side and the rest are (sometimes) likely to follow.
Obtaining this buy-in ought to be easy because such meetings are a great opportunity to make a useful point – that improved IT processes will likely make the manager’s group operate more efficiently. Any manager worth his bonus is likely to respond to that.
Among some of our colleagues, data classification exercises have a reputation for being somewhat painful. That needn’t be the case.
