The Data Readiness Checklist Every Executive Needs
Before you spend a single dollar on AI, answer these ten questions about your data. If you can't answer most of them, you're not ready.
Before you spend a single dollar on AI, answer these ten questions about your data.
The Checklist
1. Can you find your data?
This sounds basic. It isn’t. Most large organisations have data spread across dozens of systems with no clear catalogue or inventory.
2. Can you access your data?
Finding it is one thing. Getting programmatic access to it — with appropriate permissions, in a reasonable timeframe — is another challenge entirely.
3. Is your data clean?
Duplicates, missing values, inconsistent formats, outdated records. Every data set has quality issues. The question is whether you know what they are.
4. Do you know what your data means?
Is “revenue” gross or net? Is “customer” the account or the individual? Without agreed definitions, your AI will learn from confusion.
5. Is your data governed?
Who owns it? Who can change it? What are the privacy implications? If you don’t have clear answers, AI amplifies the risk.
6. Is your data timely?
If your AI model needs real-time data but your pipeline updates weekly, you have a problem no algorithm can solve.
7. Do you have enough data?
Machine learning needs volume. If you have hundreds of examples when you need thousands, the math doesn’t work.
8. Is your data representative?
Data that doesn’t reflect reality produces models that don’t work in reality. Bias in, bias out.
9. Can you connect your data?
AI gets powerful when you combine data sources. If your systems can’t talk to each other, that power stays locked away.
10. Do you have the skills to work with your data?
Data doesn’t analyse itself. If you don’t have people who can prepare, clean, and engineer your data, the AI tools sit idle.
Scoring
If you answered “yes” to 8 or more: you’re in a strong position to explore AI.
If you answered “yes” to 5-7: focus on data foundations first.
If you answered “yes” to fewer than 5: you have a data strategy problem, not an AI problem.
The Hard Truth
Most organisations score 4-6. That’s not failure — it’s just where you are. The important thing is to be honest about it and invest accordingly.