Data silos, while common, are huge sources of inefficiency in any department or organization. This problem is exacerbated in IT, where any loss of time can incur huge costs. This blog will discuss, from the perspective of NetOps and SecOps, what data silos are, why they’re bad, and what they cost.
What is a Data Silo?
A data silo is a situation wherein only one group in an organization can access a set or source of data. Data silos can result from several factors, including:
- Cultural: Competition or animosity between departments can cause those employees to keep data from each other, rather than working together.
- Structural: Especially in large organizations, data silos can stem from a hierarchy separated by many layers of management and highly specialized staff.
- Technological: Applications might not be used, or even designed, to cross-reference or add to each other. Or one department may simply not have access to a valuable app from another department because it was not purchased for their specific day-to-day tasks.
Why are Data Silos Bad?
In short, data silos cause wasted resources and inhibited productivity. There are two overarching situations that arise from data silos: multiple teams either store the same data or they store complementary, but separate, data. Each situation causes several problems.
First of all, storage costs money. Setting aside redundancy and fault-tolerance, why pay extra to store the same data in multiple places? Letting different teams access the same dataset is much more efficient. If you’re trying to find budget somewhere, this may be the first place you want to look.
Furthermore, in certain situations, the contents of those data silos will likely differ slightly. How can you determine which one is the most accurate or up to date? You also risk overwriting the current data with outdated data.
On the other hand, a team may hold data that would be useful to another team were they able to access it. NetOps and SecOps each have different tools and interfaces they use to do their jobs. But there are many situations where one could use the other’s data to solve an issue more efficiently.
For example, many security threats can be detected much more quickly by analyzing traffic data for unusual patterns. But often, SecOps doesn’t have access to that in their tools, and has to get the data by directly interacting with NetOps. This leads to critical errors, workflow gaps, and increased risk to business.
Data Silos Prevent You From Seeing the Big Picture
In 2016, F5 asked organizations how many applications they have in their portfolio. 54% of respondents said they have as many as 200 applications on their networks. 23% said as many as 500, and 15% said as many 1000. 9% of respondents even said between 1001 – 3000. Other sources point to an average of 508 applications per enterprise.
These numbers blew my mind. Even estimating NetOps and SecOps apps at a fraction of those, it quickly becomes overwhelming. Imagine trying to investigate a problem, and having to manually check each data silo and try to piece together relevant information? (Actually, many of you have probably had to.)
We created a video that explains this particular challenge in more depth:
Essentially, it slows you down not just because you have to look at each data silo one by one, but also because you have to figure out which bits of information are relevant to the problem you’re investigating.
How Much Do Data Silos Cost?
It’s hard to determine any concrete numbers without delving into a lot of hypotheticals. But we can think about it in terms of how much you value your teams’ and your organization’s time.
How much do you lose every hour that your employees can’t work because the network is down? Every day that low-and-slow data theft goes undetected? Or every month your projects and goals are pushed back?
By answering these questions, you can figure out what fixing inefficiencies is worth to your organization. And data silos are definitely a big contributor to inefficiency.
For more information on the benefits of inter-departmental collaboration, check out this blog:
