The ROI Challenge: Why Measuring Data’s Value is Hard, but Crucial

Here’s a story playing out in many data-driven organizations, and it often follows a familiar three-act structure.

Act I is Growth: A company invests in a modern data platform, a few high-value data products are built, and the business sees immediate value. Excitement is high.

Act II is the Explosion: Empowered by the platform, teams across the organization start creating more and more data products. Building is fun and fast. But beneath the surface, the costs for maintenance and infrastructure begin to quietly balloon.

This leads to Act III: The Squeeze. An alarm bell rings from the finance department. Costs are spiraling, and leadership goes into cost-cutting mode. The first instinct is to optimize infrastructure, but that only treats the symptom. The real, difficult question is: which of these dozens, or hundreds, of data products are actually delivering value, and which are just costing us money?

Squeezing the infrastructure for some sweet sweet lemon juice

Without a framework to answer this, data is treated like a technical cost center to be cut, not a strategic asset to be managed. In this post, we’ll dive into the challenges that arise from this reactive cycle and share a path forward based on our experience.

The High-Stakes Guessing Game of Prioritization

During the “Squeeze,” leaders are forced to make tough decisions with limited visibility. When every data product is presented as important, initiatives get prioritized based on intuition or internal influence rather than measurable outcomes. For the leaders making these calls, this introduces risk. A wrong bet can lead to underperformance and missed KPIs. What they need is confidence: a clear link between data investments and business returns.

Data Teams Struggling to See Their Impact

This ambiguity also affects the data teams on the ground. They are often asked to build and maintain products without ever seeing the downstream business impact. When they can’t connect their work to real-world outcomes, motivation suffers. To keep high-performing teams engaged, it’s critical that they understand how their work drives business value, not just that it keeps the lights on.

A Path Forward: Managing Data as a Business Asset

To break this reactive cycle, we need to shift our thinking from managing a technical inventory to managing a collection of business assets. This requires translating the complex data landscape into the language of business: Value, Cost, and Governance. Measuring these is notoriously hard, especially the indirect value of foundational products that support everything else. But you can’t just keep adding to infinity. A framework is needed to decide where to invest and where to divest.

Putting ROI into Practice

From our experience, building a system to manage the ROI of data products requires a few crucial steps.

Step 1: Set Clear Objectives for Value, Cost, and Governance

Before you can measure anything, you must first define what “good” looks like. The most important part of strategic leadership is setting clear, explicit objectives. This means going beyond just tracking metrics and defining actual targets for each data product.

• For Value: Is the goal to drive a specific revenue target, reduce operational costs, or achieve a certain level of adoption and usage?

• For Cost: What is the acceptable budget for this product, combining both cloud spend and the cost of people’s time?

• For Governance: What is the required level of quality, documentation, and ownership for this product to be considered “value-ready”?

Setting these objectives creates a baseline to compare against, answering the most important question for any leader or owner: “Are we on track, or do we need to course-correct?”

Step 2: Automate Measurement Against Your Objectives

Once objectives are set, the next step is to track progress against them. This cannot be a manual, quarterly exercise in spreadsheets; it must be automated and integrated. A modern approach involves a system that automatically pulls data from various sources to provide a holistic view.

This system should consolidate costs from cloud providers like Snowflake and people’s effort from tools like Jira. It should track value by measuring downstream dependencies and usage. And it should score governance by evaluating documentation, data quality tests, and ownership. This provides leaders with a high-level strategic overview while giving data product owners an actionable playbook to improve their assets.

ROI needs to have a target goal, and ideally cost and value are automatically measured

Why This is More Important Than Ever

This framework is becoming even more critical with the rise of Generative AI. GenAI will act as a massive accelerator, making it easier to create more data products, faster. While this is a huge opportunity for innovation, it also intensifies the risk of drowning in a sea of low-value, high-cost experiments.

A solid framework for setting objectives and measuring ROI is the only way to separate the high-value assets from the noise before costs spiral out of control. How would this work in practice? Learn more and reaach out to explore what it could mean for your organisation.