CIO.com published a catchy-titled article last week: "Master Data Management: Forgiveness for All Past Data-Management Sins?" The lead paragraph explained that an Aberdeen Group report found the main reason IT leaders are flocking to MDM is they believe it "can help fix the data-management mistakes they've committed over the years."
Wow. Now there's news I'd personally hate to deliver to a CEO. "Hey there, Ms. CEO. I've got some bad news and some goods news. The bad news: Mistakes were made. The good news: I can fix them for a few cool million."
I checked out the report, which is available for free via Siberian -- not surprising, given that much the piece focuses on how Siberian customers fared better in the long term than those who went it alone or who used IBM. It said the top driver for adoption was "inadequate performance of the existing data-management infrastructure."
That's not quite the same thing as saying they're using MDM to fix the mistakes they've made over the years, is it?
In the language of the CIO.com article, it sounds as if IT carelessly put together systems full of mistakes and now they're going to spend a bunch of money trying to fix them. The Aberdeen report's language, on the other hand, is less judgmental and simply says IT views the existing systems as inadequate and believes MDM can help.
Ok, so I've got your back on that one, IT. But, frankly, this report from another research firm is a bit harder to explain.
IT Pro published an article last week based on a recent survey by the Information Difference, an independent MDM research company. The survey concluded that poor data quality is costing companies tens of millions of dollars -- and despite this, a third of those surveyed had not deployed data-quality tools to stop the problem.
The articles says only 14 percent thought costs directly attributable to poor master data were less than $1 million a year. To be clear, that means everybody else thought the problem was costing more -- sometimes much more. Twenty-one percent said these data problems were costing their companies between $10 and $100 million a year.
But perhaps the most disturbing finding was more than one third said they couldn't calculate profit margins consistently across the organization because of data problems.
Andy Hayler, CEO of the Information Difference, added this commentary:
We often hear that MDM is a key concern for CIOs, yet in reality it seems there's still very little actually being done about it ... Particularly worrying is that that over half the senior IT executives we spoke to weren't consistently sure whose spreadsheet has the correct data.
That's pretty frightening. In my book, if you can't identify the correct data for profit margins, then you've got a serious business problem.
Some of this was inescapable, I suppose. We often talk about IT's infrastructure in terms of buildings and architecture, but as a reader David Wright recently pointed out, it's a bad metaphor:
That's great for skyscapers and even information systems (at a point in time); but, enterprises are not complicated, they are complex, meaning they are dynamically evolving and changing. I am not sure if architecture is the right approach for managing a non-static thing. Perhaps enterprise 'biology' is more appropriate. Life sciences may be a better analogy than architecture for an organization (organism).
I have to agree, though I doubt the two of us will be able to stop the use of the architecture metaphor. In fact, yesterday, I wrote how the need for integration should be seen as, not a series of past mistakes, but a sign of a growing, thriving IT infrastructure.
But just because you did the best you could, don't expect a parade and well-wishes when you fix mistakes of the past.
This, I think, is where IT often comes into conflict with the business. IT thinks it's doing something great for the business because it's fixing problems. But the business users are appalled to learn these "IT problems" existed in the first place. Like the CIO.com article, they're more likely to view your work as correcting the sins of the past.
I'm not sure how you can fix that, but one thing is certain: It won't help to look the other way when you know data-quality and integration issues are costing your company real money. As 12-step programs like to say, admitting you have a problem is the first step to fixing it. And maybe that's the penance IT has to pay after years of acting smarter than and separate from the business.