BI Has the Worst ROI in the Modern Data Stack – How to Escape the Service Trap and Drive Real Decisions

Published on October 27, 2025

BI Has the Worst ROI in the Modern Data Stack – How to Escape the Service Trap and Drive Real Decisions

Introduction

Business intelligence (BI) was meant to be the jewel in the crown of the data stack, the place where numbers become insights and insights become decisions. Yet if you ask most data leaders today, BI delivers the worst ROI in the stack. Teams are drowning in dashboards, executives don’t trust the numbers, and tools that haven’t evolved in 20 years are still eating up budgets.

Ollie Hughes, CEO of Count, argues that the industry has got it wrong. More dashboards aren’t the answer. AI won’t fix reporting chaos. And data teams need to stop behaving like service desks.

This article is based on Ollie’s appearance on the Data Matas podcast and translates his perspective into actionable lessons you can apply right now.

What you’ll learn:

  • Why BI tools are stuck in the past and what that means for ROI

  • How to spot the “service trap” in your team before it kills your value

  • Why accuracy alone doesn’t build trust in data

  • How ruthless prioritisation changes the credibility of a data team

  • Practical steps to create operational clarity instead of dashboard noise

Meet Ollie Hughes

Ollie Hughes is the co-founder and CEO of Count, a canvas-based BI platform built around collaboration, not dashboards. His mission is to reframe how organisations use data: not as a firehose of metrics, but as a tool for genuine decision-making.

He has spent years inside the industry, from building data teams to leading a new wave of BI innovation, and has become one of its sharpest critics. His critique isn’t abstract. It’s grounded in the frustrations most practitioners feel every day: wasted reports, confused stakeholders, and tools that don’t fit how people actually work.

“BI tools are still the same ones we were using 15, 20 years ago. They’re expensive, and all you’re really paying for is to see sales numbers sent around the company. It’s not driving decisions.”

That willingness to say the quiet part out loud, and back it with solutions, makes Ollie an important voice for data leaders rethinking their approach.

Why This Challenge Matters Now

Cloud data platforms, pipelines, and governance tools have all seen dramatic innovation. BI hasn’t. It’s still read-only dashboards that require endless interpretation elsewhere. Meanwhile, every other tool we use, from marketing platforms to document editors, has become collaborative, flexible, and iterative.

The result? Data teams are under pressure to deliver value but stuck with outdated paradigms. Many end up in what Ollie calls the “report factory”: churning out dashboards that confuse more than they clarify.

A common misconception is that more speed solves the problem. Leaders throw AI at reporting in the hope that faster answers mean better decisions. In reality, Ollie argues, the bottleneck isn’t producing numbers, it’s helping humans interpret them and agree on what action to take.

This is why now is the moment to rethink BI. As companies adopt AI at pace, trust, clarity, and prioritisation become the real levers of success.

How to Escape the Service Trap and Drive Real Decisions

1. Recognise That BI Hasn’t Evolved

Most of the modern stack has innovated, BI hasn’t. Dashboards may be prettier, and integrations with tools like dbt smoother, but the core interaction hasn’t changed.

“You read a dashboard, you discuss it somewhere else, and you try to work out what’s going on. That paradigm is still the same as 2005.”

Implementation guidance:

  • What to do first: Audit your BI output. How many dashboards exist? How many are actively used?

  • Tools/structures: Usage analytics inside your BI tool can show adoption and engagement.

  • Watch-outs: Don’t assume integration features equal innovation, the form factor is what matters.

  • Expected benefit: Clear view of which reports genuinely support decision-making.

2. Avoid the Service Trap

Many data teams confuse activity with impact. They answer every request, build dashboards for every stakeholder, and believe they’re adding value. In reality, they’re generating information overload.

“Just doing what the business asks of you floods the company with chaos. If the business is asking stupid questions, the data team is going to be producing stupid answers.”

Implementation guidance:

  • What to do first: Count the number of dashboards per employee. If it’s high, you’re in the trap.

  • Tools/structures: Introduce a request filter or impact sizing model.

  • Watch-outs: Saying “yes” to everything positions your team as a service desk, not a strategic partner.

  • Expected benefit: Fewer but higher-value outputs, stronger alignment with the business.

3. Build Trust Beyond Accuracy

Most data leaders obsess over accuracy. But Ollie warns: accuracy alone doesn’t create trust.

Imagine telling the CEO: “Our regression says move all marketing spend from Channel A to Channel B.” Even if it’s correct, they won’t act on it unless they understand how you got there.

“Trust comes from methodology, transparency, and track record, not just from being right.”

Implementation guidance:

  • What to do first: Make “show your working” a standard practice for every analysis.

  • Tools/structures: Adopt lightweight documentation or visual lineage tools to explain methodology.

  • Watch-outs: Overcomplicating explanations can backfire, clarity beats detail.

  • Expected benefit: Executives who feel confident enough in the process to act on insights.

4. Prioritise What Really Matters

Ollie’s strongest advice: not all requests are equal. Some will change the trajectory of the business; most won’t.

“If you’ve solved the most important problem the business has today and the CEO recognises that, you’ll be remembered for it. That’s what matters.”

Implementation guidance:

  • What to do first: Track where your team’s time goes: maintenance vs problem-solving.

  • Tools/structures: Create a simple payroll allocation dashboard – % time on top 3 business priorities.

  • Watch-outs: Saying “no” is hard, but without it your team’s value will always be capped.

  • Expected benefit: Senior leaders see the team as solving the biggest problems, not just keeping the lights on.

5. Create Operational Clarity, Not More Dashboards

In a world where every SaaS product spits out metrics, the role of the data team is to simplify, not add noise. Ollie calls this “operational clarity.”

“The job is to show the forest, not just the branches. Visualise the business, make it feel simple, align everyone on what matters.”

Implementation guidance:

  • What to do first: Build a single-page growth model that shows how key metrics relate.

  • Tools/structures: Collaborative BI tools (like canvas-based environments) that allow business and data teams to work together.

  • Watch-outs: Don’t replicate existing reports, focus on connections, not duplication.

  • Expected benefit: A business that understands itself better, asks better questions, and makes clearer decisions.

Putting it All Together

The path forward isn’t more dashboards or faster charts. It’s about shifting from outputs to outcomes.

A realistic sequence:

  1. Audit existing dashboards and usage

  2. Filter requests through impact sizing

  3. Make transparency part of delivery

  4. Redirect at least 50% of team capacity to the biggest problems

  5. Replace scattered reporting with a unifying model of the business

Signals of success: fewer but more impactful outputs, leaders asking sharper questions, and a measurable increase in trust in data-driven decisions.

Real-World Impact

Count’s customers have already applied this model. By shifting away from dashboard churn, they’ve reduced noise, improved decision-making speed, and redefined how business and data teams work together.

The result isn’t just cost savings. It’s a cultural shift: data teams that no longer see themselves as report writers, but as strategic partners shaping the direction of the business.

Your Next Move

The lesson from Ollie Hughes is simple: stop measuring success by the number of dashboards you ship. Measure it by the clarity and decisions you enable.

Focus your team’s time on the most important problems, show your working, and embrace operational clarity. That’s how data leaders can turn BI from the lowest ROI into one of the highest.

🎙️ Listen to the full conversation with Ollie Hughes on the Data Matas podcast for more actionable insights.

Dig Deeper

#Blog #ai #Data Analytics Strategy #Data Engineering #Modern Data Stacks #POWERBI

Seen a strategy that resonates with you?

BOOK A CONVERSATION