Working on a data migration project gives you a unique opportunity to learn where your organization has fallen short in its data management strategy. It's when you start to explore your legacy data landscape that you get a feel for how big a silo challenge your company has.

It wasn't always this way, of course.

In the 1960's, organizations mostly relied on one centralized system to manage the computing function. It may have required a room the size of your house, but having all your eggs in one basket was the order of the day. Let's call these earliest beginnings Silo 1.0. After all, in most organizations, there was only the one silo.

In the 1980's came the decentralized model of computing. Suddenly it was (relatively) easy and more affordable to scatter new systems and data stores around the enterprise in a client-server arrangement. Users began to enjoy the freedom of working on their fat clients to create 'fiefdoms' of locally sourced data.

Silo 2.0 was born.

In the 2000's we saw the arrival of higher performance Internet connections and powerful cloud computing infrastructure. A smorgasbord of lower-cost data landscapes and business functionality was now available, and the third generation of computing emerged – Silo 3.0.

Which brings us to today.

Self-service data provisioning. Real-time analytics. BYOD. Big data. The list of data demands goes on, and the challenge for modern data leaders has never been greater. So what can history teach them?

Silo proliferation can never be eliminated. Data leaders need to accept that just over the horizon will be a new wave of silos. Further, I believe it's inevitable that users and departments will continually try to source their data and apps, often independently of any central IT strategy. And finally, we just need to get smarter at managing data beyond the conventional boundaries. There are several reasons why:

• First, the organizational boundaries of our data landscape are blurred. Company data can reside in a cloud-based center well outside of the organization. We may have conceptual ownership but physical ownership is often perceived as a costly overhead, best avoided.
• Second, the boundary between business and IT is changing as the business demands more and more agility. Gone are the days of the IT team working year after year to develop new systems. We're far more likely to outsource our applications, buy off the shelf or lease as a service in the cloud.
• Third, the boundary between users and their data has broken down. Organizations are increasingly adopting self-service methods of data analytics and access. The days of waiting a month for your custom report are over. Speed and agility are the order of the day, and we have far more access to information than we used to.

Clearly the old approach of totally collapsing silos and forcing functionality on users doesn't work. If the data and its associated applications don't fit their needs, users will create alternatives. So how can data leaders cope with this new landscape they've inherited?

I think data leaders need to focus on several areas.

We must roll back the clock and try to regain control of information architecture and the core techniques involved. For example, it's frightening how many companies today pay lip-service to disciplines such as conceptual and logical modeling. Conceptual modeling forces the organization to consider the "to-be" business models their information has to support. It is these models that reflect the roadmap to a future state.

Lack of vision and coordinated strategy are why most organizations are always playing catch-up with their information management. We have lost sight of this and many other disciplines, focusing too much on IT architecture; it's not the same animal.

The emergence of the chief data office (CDO) role should level the playing field, provided the CDO reports to the CEO and can wrestle control for information architecture away from IT.

Finally, if traditional boundaries are breaking down, then we desperately need to improve data governance and stewardship. Silos need governance and careful data design to control alignment with today's information architecture, with a view on enabling a road map to the future.

Silos and shifting boundaries are here to stay. Data leaders need to adapt sooner rather than later.