Don't Establish Data Management Standards

A recent survey of Enterprise Architects showed a lack of standards for data management.* Best practices has always been about the creation of standards for IT, which would lead us to think that lack of standards for data management is a gap.

Not so fast.

Standards can help control cost. Standards can help reduce complexity. But, in an age when a data management architecture needs to flex and meet the business need for agility, standards are a barrier. The emphasis on standards is what keeps IT in a mode of constant foundation building, playing the role of deli counter, and focused on cost management.

In contrast, when companies throw off the straight jacket of data management standards the are no longer challenged by the foundation. These organizations are challenged by ceilings. Top performing organizations, those that have had annual growth above 15%, are working to keep the dam open and letting more data in and managing more variety. They are pushing the envelope on the technology that is available.

Think about this. Overall, organizations have made similar data management technology purchases. What has separated top performers from the rest of organizations is by not being constrained. Top performers maximize and master the technology they invest in. They are now better positioned to do more, expand their architecture, and ultimately grow data value. For big data, they have or are getting ready to step out of the sandbox. Other organizations have not seen enough value to invest more. They are in the sand trap.

Standards can help structure decisions and strategy, but they should never be barriers to innovation.


*203 Enterprise Architecture Professionals, State of Enterprise Architecture Global Survey Month,2012

**Top performer organization analysis based on data from Forrsights Strategy Spotlight BI And Big Data, Q4 2012


Demolish the dams because we are too slow to modify them?

I agree with the final conclusion "Standards can help structure decisions and strategy, but they should never be barriers to innovation."

I disagree with "...standards are a barrier".
I supposed more of us would agree that standards are not carved in stone, are subject to review and change thus the real barrier is the inflexibility in adapting standards to new requirements.
If you have a cumbersome process to modify standards (or anything else for that matter) when required then blame those processes and not the standard (which is just a document, a manifestation of your thinking, values etc.).
If you have such problems with standards (modifying them when required), I bet you generally have problems in implementing necessary business changes in general i.e. its just one symptom of a much larger problem.
If you need to be agile you need to practice it holistically and not selectively. Getting rid of standards for the sake of being agile because you are yourself not agile enough to start with to make changes does not sound right.

"...standards are a barrier"

I think you articulated this well and show that common ground is needed on what we mean when talking about standards. Standards can architectural, procedural, and categorical.

Architecture: vendor standard, platform standard, delivery standard, design standards
Categorical: business rules, taxonomy, models, metadata, etc.
Procedural: processes, operations, development, etc.

You are correct to say that procedural is what most often gets in the way. It is a change management challenge. But, I think there is also the challenge that skills are a factor.

You mention, "getting rid of standards for the sake of being agile...does not sound right". I would say that agile is a standard or at least a core principle.

Would it not be better to define a standard in a manner that encourages evaluation and evolution rather than setting something in stone? Have we defined a framework for standards too black and white so that the effort to be agile, be flexible, and change is inhibited? It could be that standards are a dynamic continuum of principles and policies rather than definition.

Depends what you mean by "standards"

If you mean "we are only going to use Product XYZ for data management" then I can see why this might be a barrier.

But lack of a data management standard for what constitutes valid data for a customer, or a product - that is something that you ought to have defined as a standard, otherwise your consuming business processes are going to suffer, as well as analytical processes that depend on it.

No control leads to caos (not talking about cost)

Standards are created to control the way data is shaped and managed across the organisation. Without standards you would be calling the same entity with different names, different definitions, you would not be able to reuse structures, etc.; among with many other problems. I love to let more data in for improved decision making, but in a controlled way. Data Integration is one of the best ways of getting value out the resulting information, and this is enabled by rules and standards that let this integration to happen.
I would be amused to watch the reaction of an ISO engineer while reading this article :)