Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Why your data strategy must evolve

Instead of storing everything forever, businesses need a documented and thoughtful data plan.

[Photo: Getty Images]

Profile photo of Chongwei Chen

President and CEO · DataNumen, Inc.

Driven by a rapid growth in social media usage and a recent surge in AI systems, global data creation is expanding at an alarming rate. Statista predicted that 181 zettabytes of data would be created in 2025 alone. Organizations are also quickly discovering that the amount of data they need to store is already stretching their budgets. The traditional approach of “store everything forever” is not practical or feasible.

This unchecked data growth has led to data sprawl, where data gets spread across different locations, formats, and systems. At times, organizations have limited control over this unmanaged data storage, and it comes with huge costs and compliance risks. The infamous Equifax data breach in 2017, where at least 40% of Americans had their data compromised, can be directly attributed to data sprawl. A loosely managed consumer complaints portal allowed hackers to get into the company’s systems and access millions of data records.

Data sprawl has forced organizations to invest in more data security resources, and is drastically inflating data backup costs. The situation is rapidly becoming untenable, and organizations are shifting to data lifecycle management, which treats data as a dynamic asset.

THE UNSEEN COSTS OF DATA HOARDING

Typically, costs associated with data storage are broadly confined to hardware costs and cloud subscriptions. However, as uncontrolled amounts of data are stored in your systems, you can be responsible for backup tool licence costs, manpower costs associated for cataloguing and securing data in different systems, and even energy costs for running data centers.

There can also be an impact on recovery outcomes. When you have unmanaged data across different systems, recovery becomes a daunting task. Imagine a cyberattack that forces you to get back high volumes of data, much of which is redundant or low value. Until everything returns, your core systems may remain unavailable, and this impacts recovery timelines.

Compliance can become a nightmare, too. When you have data accumulating in different systems at a brisk rate, you may run afoul of regulations like GDPR or CCPA. Many of these regulations have mandates that force companies to delete data they do not use. It’s not an excuse if your team is unaware of data items stored in an obscure system. Unnecessary and loosely managed data drastically increases your chances of a data breach and resulting penalties.

CLASSIFICATION, TIERING, AND AUTOMATION

Unplanned data growth has transformed data storage into a sprawling liability. We need to act systematically to weed out the clutter. Here’s how to do that.

Build a Value Index: The first step of effective data management revolves around understanding your data and its relative value to the organization. Organizations should implement a value index broadly classifying data based on its business value, compliance requirements, and access frequency. In real-world terms, they can be labelled as:

Mission critical: Active transaction databases needed for day-to-day operations.

Transactional: Operational data needed for intermittent use, but a notch lower on priority than mission-critical data.

Compliance archive: Data mandated by law for retention irrespective of business value.

Cold or junk data: Obsolete content or irrelevant data that serves no business purpose; retention is not covered under any regulatory mandates.

Enable tier storage for cost control: Once you have properly classified the data, you enable the foundation to implement a tiered storage process, which can drastically lower costs. Data can now be stored on different media types based on performance needs and access frequency.

Hot storage: Consists of SSD and other high-performance media to store mission-critical data requiring immediate access.

Warm storage: Usually uses standard disks to store transactional data.

Cold storage: Can be typically be stored on the cloud to maintain compliance records and other historical data.

Take advantage of AI/ML tools: With the advent of AI/ML tools, it has become convenient to deal with unstructured data that often forms the bulk of an organization’s data. These advanced tools can quickly classify data, find personally identifiable information and other sensitive data, and it can come in handy for content tagging.

ACTIVE RETENTION AND STRATEGIC DELETION

Once data classification has brought order to your organization’s data, you need to implement governance for its longevity.

Review the retention policy: Every data asset in storage has a cost. Hence, your data governance rules should have retention timeframes and even disposal procedures. Based on the data category, lay out a specific time frame for data retention. Then clearly document policies for data disposal; don’t just focus on deletion triggers.

Embrace defensible disposal: Next, implement defensible disposal—an auditable deletion paradigm that can stand up to any legal scrutiny. Log files should be duly maintained, clearly showing what was removed from the system and when. Besides protecting you from legal hassles, defensible disposal reduces the overall attack surface and minimizes potential cyberattack damages.

Optimize business continuity and recovery plans: In this new reality, targeted business continuity plans focused on critical data have become the norm. Reduced and classified data volumes facilitate quicker backups and allow you to enable faster recovery of mission-critical data.

THE LEAN DATA ADVANTAGE

The concept of efficient data management relies less on saving disk space and more on building a lean, resilient, compliant business model. When we reduce waste, regulatory exposures, and increase operational efficiency, we also unlock strategic value. The future’s technology leaders will be judged not on the size of their ginormous data lakes, but by the efficiency of their data pipelines and ability to reduce clutter.

Chongwei Chen is president and CEO of DataNumen.