How do I convince business people that managing data costs money?
June 13, 2012 2:30 PM   Subscribe

Help me with some IT management philosophy. How do I convince people internally that managing data costs money? Gory details inside

I work in a business that is very information/data intensive.  Most of the business is managed in one large database, large both in width (thousands of columns) and length (millions of rows).  Over the years, people on the business side inevitably ask to increase the size and complexity of their data needs with little resistance from our IT department.  What then happens is you end up with monsterous amounts of data, most of which is poorly defined, lacks adequate controls, and is often times not used or assumed to be incorrect anyway.  In other words, the “cost” of properly maintaining this data is not adequately weighed against the “benefits” in additional revenue to the business.
I’m not in IT, so I don’t have an academic background on this topic, but it strikes me as something that must commonly be dealt with in all sorts of industries.  I want to raise this topic internally and have philosophical discussions about how to prevent it.  I’m looking for tools to use to facilitate: research reports, websites, academic studies, best practices… something to help quantify to business people why simplicity is a virtue and/or suggestions on how to solve this issue
Related background: the software using this data also requires continuous development (as our business model changes often) and the software itself also suffers from excessive complexity, which makes subsequent changes more and more difficult over time
posted by bpdavis to Computers & Internet (4 answers total) 4 users marked this as a favorite
While there is a great deal of material addressing this from an IT perspective, it sounds like there is a more fundamental management issue at play here.

How does the business handle cost-benefit analysis in general?

Until you address that fundamental mangrment issue, no one in IT or in the rest of the business is going to want to do the hard work of changing things.

Once you address that, you can start getting into the question of how best to estimate costs and benefits of IT development and operations.
posted by Good Brain at 3:25 PM on June 13, 2012

I need to round that comment a little more. I see three major pieces
1. Establish the need for a culture of analysis and accountability in the business as a whole.
2. IT develops a culture of estimating costs, risks and benefits for their development and operations.
3. IT uses this as a foundation of their relationship with the rest of the business.

This is going to be an iterative process, but it needs to start somewhere, and before long, it needs strong support from most or all of the players in order to come to full fruition. Further, it needs it needs to aspire to cooperation, collaboration communication and negotiation, rather than animosity and rivalry.
posted by Good Brain at 3:38 PM on June 13, 2012

You need a good DBA or data architect with experience in data vaulting... someone who's job is to manage your business data's size and complexity. Spending money in the short term will save a lot of it in the long term.

But first, you need to convince IT there's a problem. Define a set of metrics for your business processes that relies on access to the data and focusses on your specific pain points - the time it takes to complete a task in man-hours, the number of errors in the process, how often a task needs to be redone, the cost to run the process, etc., and measure it week to week and month to month to show degradation, or the increase of the cost to support the data vs. the revenue generated by the business processes. IT wonks looooove numbers like this, no lie - it really helps them in their job.
posted by Slap*Happy at 5:39 AM on June 14, 2012

Even an inexperienced hack that hasn't studied Data Model Patterns by David Hay or any other data modeling book from the last ten years knows better than to put everything in one big table. It's 2012—who treats a modern RDBMS like COBOL flat files anymore? (Besides Lawson, of course. They call it being "database neutral", but it means the same thing.)

It sounds to me like someone high-up in your company is proverbially cutting the ends off the roast. As such, this isn't really a technology problem, it's a leadership problem. Especially if the leadership is of the old-school, IT is treated as an expense instead of a capital improvement. Here in the 21st century, especially for businesses that sell products or services, your IT infrastructure and the software that runs on it contributes as much as your physical plant to your success.

Is is there enough revenue to justify the expense of doing it right? If you had a better system, could you expand and start making enough to justify it? Any decent consulting outfit could take what you have and turn it into a bespoke solution including a normalized database and maintainable software. The time and cost would vary depending on the size and complexity, but I think it would be very competitive with implementing a mid-range ERP.
posted by ob1quixote at 10:20 AM on June 14, 2012

« Older I've saved up $100k in my first year of work. I...   |   The sum of the square roots of any two sides of an... Newer »
This thread is closed to new comments.