If AI Is So Smart, Why Are We Still Struggling for ROI?

Are you aiming to use AI to improve your bottom line? There’s no shortage of AI frameworks, roadmaps, and best practices for building successful data, analytics, and AI initiatives. From capability maturity models to AI readiness checklists, you can find a how-to guide for just about every step of the journey.

So why are so many organizations still failing to see meaningful return on their AI investments?

In the coming weeks, we’ll share eight of the most common ways organizations lose ROI in their data, analytics, and AI initiatives. But rather than a checklist or blame game, each post will center around a question—the kind that leaders need to ask early and often. For example:

• Do our weekly insights clearly advance our top business goals?
• Do our teams have the right mix of skills, tools, experience, and data to deliver business results?
• Is AI driving us into danger because we don’t understand its blind spots?

These questions aren’t theoretical. They’re practical, diagnostic, and often revealing. If you’ve ever wondered why your data investments aren’t delivering as promised, this series is for you.

Many organizations are still grappling with achieving ROI from traditional analytics, and AI only adds complexity. With less clarity around how AI reaches conclusions (interpretability), demands for new data types, workflows, risk management, and outputs, ROI can seem almost unattainable.

Treating AI as a plug-and-play extension of an already struggling analytics program only widens the gap to achieving ROI and amplifies organizational frustration. That’s what one organization we worked with experienced.

This global consumer packaged goods company had invested millions in modernizing its tech stack and building predictive models. On paper, they were checking all the boxes: new tools, new talent, executive sponsorship.

But insights from models weren’t getting used. Why? The sales and supply chain teams didn’t trust the outputs, and leadership hadn’t aligned on how success would be measured. The tech was sound, but the business context was missing, so ROI never showed up.

We set out to build a new framework for ensuring ROI but quickly realized the world didn’t need another framework. Organizations need clarity on why the existing practices often don’t deliver.

Organizations often latch onto a framework thinking it will fast-track success, but frameworks are just tools. They don’t diagnose your specific blind spots, risk tolerances, or business context. They don’t tell you whether your data priorities align with your top business objectives or whether your team even knows precisely what good looks like.

We’ve seen companies pour resources into platforms and people only to realize—all too late—that they never agreed on how value would be defined and measured.

Meanwhile, analysts are busy answering ad hoc requests, but no one is sure whether those answers are moving the needle. Dashboards pile up, but key decisions are still made on gut feel because the data isn’t trusted, timely, or maybe even accessible.

Too often, analytics produces activity without clarity. Teams spin up different versions of the truth, insights get stuck in slide decks, and new tools get deployed without a clear purpose. AI adds even more urgency to these challenges.

If your business isn’t aligned on how to prioritize analytics, adopt insights, or measure progress, then layering AI on top won’t help. It will only deepen the cracks.

To help make sense of where and why ROI is being lost, we developed a flexible model in a simple structure that keeps the focus on what matters. It centers on four key dimensions every data and AI initiative should align to:

1. Business Strategy:

• Do our leaders have the data literacy to ask the right questions and propose the right projects?
• Is AI driving us into danger because we don’t understand its blind spots?

2. Resource Allocation:

• Do we prioritize analytics projects according to the business value they deliver?
• Are we investing in solutions that measurably move the business forward?

3. Project Execution:

• Where are we still making gut decisions because the data isn’t reliable, accessible, or timely?
• Do our teams have the right mix of skills, tools, experience, and data to deliver business results?

4. Strategic Deliverables:

• Do our analytics teams create and employ a single source of truth that keeps the business aligned?
• Do our weekly insights clearly advance our top business goals?

This all wraps around one central question: Are we seeing impact?

In this series, we’ll arm you with the questions most leaders never think to ask so you can expose the hidden gaps draining your AI and analytics ROI and shut them down for good.

Instead of another framework, we are offering a way to uncover where ROI is being lost and get it back. This isn’t a technical deep dive; it’s a strategic reflection for leaders trying to get more out of what they’ve already built.

Whether you’re in the middle of a transformation or just starting to scale, the questions we’ll dive into are designed to surface risk, challenge assumptions, and refocus your efforts on value.

It all starts with asking better questions.