“Educate your business stakeholders on the data lifecycle”

JAXenter: You recently completed a study on enterprise data management trends. What surprised you most about the study’s findings?

Krishna Subramanian: We know data growth is exploding, as are its costs, but what was surprising is that our survey found that IT managers are thinking more strategically about data management. While cost-cutting is always a top priority, migrating more data to the cloud and having better analytics on their data ranked higher. This shows that businesses view data as an extremely valuable asset and see data management as a way to drive competitive advantage.

SEE ALSO: “Data storage cannot be a set and forget exercise”

JAXenter: In the press release you issued about the study, you talk about the scale of data companies are managing, the cost to do so and the value their data represents. You also mention that analytics can help with planning and cost savings. How does that work?

Nowadays there are so many reasons to move data to the cloud – lower operating costs and complexity is one reason, but using the elastic compute in the cloud for new applications like AI and Big Data analytics is another reason.

Krishna Subramanian: At Komprise, we like to talk about “know first” in data management. People often think the cost of data is the cost of storage and so if they have more data, they should just buy more storage at a lower cost. But 80% of the cost of data is not in the storage – the bulk of the cost is in the active management of the data. That’s why the endless cycle of buying more storage is not sustainable. Instead, by understanding what your users are doing with data – how often it is accessed and by which groups, its priority in daily business operations, where it is stored and the cost to store and back it up – you can make the right decisions about how to manage it.

As an example, with our customer base we find that 70 to 80% of data is cold or infrequently accessed. Yet most of it is living on expensive storage and backups when by moving it to lower-cost durable storage — be that in a private or public cloud – you can cut up to 75% on storage costs. That’s millions of dollars a year in a large enterprise which can be spent on much more strategic initiatives that are critical to customer retention. And analytics is key to having these conversations with users, and transparent data movement means you don’t have to ask users to look for data elsewhere or change their behavior.

JAXenter: The survey says that the top priority for IT leaders is moving data to the cloud, but it also says half of companies are still storing data on-premises. Under what circumstances should a company move its data to the cloud? Is there still a good argument to be made for on-premises storage?

Krishna Subramanian: Nowadays there are so many reasons to move data to the cloud – lower operating costs and complexity is one reason, but using the elastic compute in the cloud for new applications like AI and Big Data analytics is another reason. A key concern for enterprises moving file data to the cloud is that their current applications and users should not have to change to use the cloud. This is why many IT leaders are using a hybrid cloud approach where hot data stays on-premises and cold data is transparently moved to the cloud such that it can still be accessed by users as before. Companies like Pfizer are tiering cold data to the cloud so they can not only reduce costs but also use the data in the cloud natively for data lakes and Big Data Analytics.

Many IT leaders don’t want all their assets in the cloud – much less one cloud. So multi-cloud and hybrid environments give IT organizations the ultimate flexibility to match workloads to the best possible environment across the key parameters of cost, performance, risk management and user/marketplace requirements. Of course, that makes everything much more complex for IT to manage– but that’s another story!

JAXenter: Let’s go back to the subject of data visibility. What are some examples of data management decisions companies need to make that can be improved with analytics?

Krishna Subramanian: There are many untapped ways to use analytics to improve data management – and by default help the business and users in many ways. We are just starting to learn about these. For instance, consider compliance. When employees leave the company, their data often lurks around and could create compliance issues. Data management tools can find such zombie data, wherever it’s stored, even in the cloud, and enable managers to make the best decisions about whether to move them to archives or delete them altogether.

In another example, ransomware is now afflicting companies in all industries and in all sizes of organizations. With visibility into data access trends – for instance if you suddenly see a high number of downloads from a file share that is usually pretty quiet – you may want to investigate and find out if there’s a problem. Sharing these data trends with your cybersecurity team is another layer of defense against hackers who are now targeting storage appliances and backup software as well as application servers.

SEE ALSO: Five Data Models for IoT: Managing the Latest IoT Events Based on a State in Apache Cassandra

Analyze data and storage assets thoroughly and have a system for doing this continuously so you always understand your data footprint.

JAXenter: One more question: If the end goal of data management is to cut storage costs and create new value from unstructured data, what are 2-3 steps IT and storage leaders can take today that will help them achieve this goal eventually?

Krishna Subramanian:

1. Analyze data and storage assets thoroughly and have a system for doing this continuously so you always understand your data footprint.
2. Educate your business stakeholders on the data lifecycle. What that means for your organization and how you will continue to protect data and also make it easily accessible and usable by departments when required. Data management is fluid – not a set and forget exercise. And this requires new thinking and awareness.
3. Find the right partners. While that sounds like the typical thing that a vendor would say, data management and the overall technology landscape is changing too quickly to have all the knowledge in-house. Select a few key partners across data science, data analytics, storage, cybersecurity, and infrastructure to help guide your way and give you the support in the form of services and tools that can automate data management processes like data archiving and deliver the ultimate savings to the organization. The end goal is to create a flexible, reliable data management foundation that puts the right data in the right place at the right time and enables employees to uncover new insights from data which hopefully, will deliver value to your end customers.