Data silos are separate islands of information. They help departments keep things private, but can hinder an organization’s ability to make data-driven decisions as a whole.
Why so many organizations delay data consolidation
You can’t consolidate what you don’t know exists. In many mid-size and large organizations, they don’t even know all of the data silos that exist.
The value of the data is also disproportionately held by the owners, creators, and most frequent users of the data. They know exactly what it is, where it is, and what they can do with it.
A consolidation effort, no matter how well designed or executed, is change — and change can have an immediate and negative impact on someone’s ability to do their job. If they can’t do their job effectively, they are at risk. And a data silo consolidation project is a risk aiming at the heart of their job.
When to retain data silos
There are always certain datasets within an organization that need to remain siloed for legal/compliance reasons.
Sometimes, innovation-focused companies even prefer data silos if setup is faster than integrating with existing data stores.
Other times, you may set something up that doesn’t end up becoming production. If your data is consolidated, you’ve likely made changes to enterprise data stores that now have to be torn down. Siloed data sidesteps these scenarios.
Silos are the reality whether they’re a best practice or not. As long as you have well-documented schemas and APIs to access siloed data, your organization will still be able to make insightful decisions.
How to convince your boss that data silo consolidation is a good idea
As soon as stakeholders understand that access permissions don’t change automatically as part of this project, they’re far more likely to support it. Emphasize that access to data that helps with decision processes can be granted easily, and that permissions can be properly controlled.
More importantly, explain how people are currently making decisions using less-than-perfect data simply because all of the data isn’t there to inform their decisions in the first place.
What to avoid when consolidating data silos
What happens is the organization sees a way to house a tremendous amount of data very cost-effectively (via the actual storage cost), but doesn’t put any consideration into the costs associated with actually accessing the data.
What ends up happening is that the costs increase exponentially.
When the chargebacks come from IT to each group for accessing their own data, they spin up their own resources to house it because it’s cheaper and easier to do it themselves…and they’re right back to silos.
One last bit of advice
What we see happening with data silos is the same thing we see with with cloud. Organizations are racing to achieve a technical objective without stopping to make sure the strategy aligns with their business objectives.
It’s so cheap and easy to get resources, that people act too quickly.
The cost of unwinding or correcting these quick actions is high and rarely factored into initial decisions.