There’s a popular image of Irish people as talkers, storytellers, and professional charmers.

There’s also a popular image of data and AI people as quiet, analytical, and emotionally attached to spreadsheets.

I like to think I’m a little of both.

One of the things we say back in Ireland is ‘don’t let the truth get in the way of a good story’. When it’s about banter and entertainment, taking liberties is an important way to keep your audience engaged and add value. 

Taking liberties with our data stories doesn’t quite work the same way. Our data stories need to use reliable historical sources that stand up to scrutiny. The data must be factual, not fiction.

And while the data must be trustworthy, it’s still important that it tells a story we can understand and inspires us to take action. The data will depress us at times and elate us at others, but the most important thing, whatever the emotion, is that we can trust the data. Facts drive action, and action gains traction.

The Explanation Rule

Can you explain your data model in plain language?

If the explanation relies on “it’s complicated,” it might be risk dressed up as intelligence.

Good data and AI work should be understandable. Not oversimplified - just clear.

Because ultimately, the goal isn’t insight. It’s action.

What This Looks Like in Practice

If data is meant to drive action, and be trusted while doing it, then a few things start to matter a lot more.

12 Things to Consider for High-Impact Data:

1. Design for decisions, not for decoration.
   Every dashboard should answer a specific business question and drive a specific action.

2. Start with the outcome in mind.
   Ask: What behavior should change because of this dashboard? If nothing changes, it’s noise.

3. Focus on a few critical metrics.
   Clarity drives confidence. Too many KPIs dilute attention and paralyze action.

4. Make trends and comparisons obvious.
   Performance only matters in the context of targets, prior periods, or benchmarks.

5. Tailor dashboards to the audience.
   Executives need direction. Managers want levers. Analysts seek depth.

6. Make it easy to drill down.
   A great dashboard surfaces the signal but allows deeper exploration when needed.

7. Align metrics with strategy.
   If your dashboard isn’t aligned with company priorities, it’s just reporting, not intelligence.

8. “Museum dashboards.”
   If no one interacts with it or uses it in meetings, it’s a wall decoration.

9. Vanity metrics.
   Measure what drives growth, profitability, and operational excellence — not what looks impressive.

10. Complexity.
    If it takes five minutes to understand what’s happening, it’s too complicated.

11. Data quality and governance.
    Beautiful visuals built on unreliable data destroy trust instantly.

12. Ownership.
    Every dashboard should have a clear owner responsible for accuracy and relevance.

Final Thought

I often say I’m not the smartest person in the room.

That’s not false humility. It’s just practical. In data and AI, there’s always someone who knows more about the model, the system, or the math.

My job is to keep asking better questions:

What action is this data telling me to take? How does this actually add value?

Activator is one of my top strengths. I need things to do - preferably the right things.

Data gives me that direction. And the more I trust it, the more confidently I can move.