By now, we’ve talked about why so many AI initiatives struggle.
- Poor ownership.
- Weak governance.
- Conflicting definitions.
- Undocumented business logic.
- Missing metadata.
The good news?
You don’t have to solve all of those problems before you begin preparing for AI.
In fact, trying to solve everything at once is one of the fastest ways to make no progress at all.
The organizations that succeed don’t build AI readiness overnight.
They build momentum.
Here’s how I’d spend the first 90 days.
Days 1–30: Understand What You Already Have
The first month isn’t about buying technology.
It’s about taking inventory.
Start asking questions you’ve probably never documented.
- What are our most important datasets?
- Who owns them?
- Which reports drive executive decisions?
- Where are our critical business definitions documented?
- What systems feed our warehouse?
- Where do we already have metadata?
Don’t worry if the answers aren’t perfect.
Your goal isn’t perfection.
Your goal is visibility.
You can’t improve what you can’t see.
Days 31–60: Build Trust
Now that you understand the landscape, start improving confidence.
Pick a handful of business-critical data assets.
Not hundreds.
Just five or ten.
Work with the business to answer questions like:
- Who owns this?
- What does it mean?
- How is it calculated?
- Where does it come from?
- Who approves changes?
- What downstream reports depend on it?
This is where trust starts.
Not because the data changed.
Because everyone finally understands it the same way.
Days 61–90: Operationalize What You’ve Learned
This is where many organizations stop.
Please don’t.
Take what you’ve documented and make it part of the platform.
Connect ownership to your catalog.
Capture lineage automatically where possible.
Associate business definitions with your datasets.
Review access based on ownership.
Treat metadata as something the platform uses—not something people occasionally update.
This is where governance begins shifting from documentation to operations.
Start Small
One of the biggest mistakes organizations make is trying to make the entire enterprise AI-ready.
That’s rarely necessary.
Choose one domain.
Finance.
Sales.
Customer.
Operations.
Build one area exceptionally well.
Learn from it.
Refine the process.
Then expand.
Success creates momentum.
Momentum creates adoption.
Adoption creates cultural change.
Technology Isn’t the Hard Part
Notice what isn’t on this 90-day plan.
No model selection.
No GPU discussions.
No prompt engineering workshops.
No vendor bake-offs.
Those conversations will come.
But they’ll be far more successful when they’re built on a foundation of understanding rather than assumption.
The Human Side
The first 90 days are really about something much bigger than technology.
They’re about conversations.
Getting Finance and Sales to agree on definitions.
Helping business owners understand why lineage matters.
Giving governance a practical purpose.
Turning metadata into something people actually value.
Preparing for AI isn’t just a technical transformation.
It’s an organizational one.
Final Thoughts
Organizations often ask how long it takes to become AI-ready.
The honest answer is:
Probably years.
But becoming more AI-ready?
That can start this week.
You don’t need to rebuild your platform.
You don’t need to replace your warehouse.
You don’t need to buy another AI product.
You simply need to begin understanding your data better than you did yesterday.
Because AI readiness isn’t a destination.
It’s a discipline.
And like every worthwhile discipline, it begins with the first step.
Leave a Reply
You must be logged in to post a comment.