I’m seeing companies invest crores in AI platforms for analytics. Most of them might just fail if they don’t address one thing: their data foundation.
Let me explain this.
The typical pattern
Sales defines “active customer” one way. Finance defines it differently. Product has a third definition. Customer data lives in four systems, each telling a different story.
Then the company turns on AI to analyse customer behaviour. It gives contradictory insights. One query says churn is up 15%. Another says it’s down 8%.
Leadership loses confidence. The project gets shelved.
Here’s what nobody wants to hear: AI platforms can’t fix bad data. They just expose it faster.
A pattern I keep seeing
After 15 years in data infrastructure, I keep seeing the same sequence play out:
Leadership invests in AI-powered analytics. AI surfaces inconsistencies nobody knew existed. Teams scramble to explain why the numbers contradict each other. The initiative loses momentum.
What the companies who succeed do differently
The few companies actually succeeding with AI for analytics didn’t start with the platform. They started with the boring, grunt work:
- Aligned on what metrics actually mean
- Built visibility into data lineage
- Cleaned up duplicate logic
- Made data quality a priority, not a backlog item
Yes, it’s not exciting work. Nobody writes Medium posts about “how we standardised our revenue definition.”
But it’s the difference between AI that delivers insights you trust and AI that becomes an expensive line item in your Annual Operating Plan.
The real question
So if you’re investing in AI for data and analytics, the question isn’t “which platform should we use?”
It’s “do our teams even agree on what we’re measuring?”
I’ll be using this space to share more on AI platforms for analytics and insights, and how this landscape is evolving. If there’s a specific topic you’d like me to cover, let me know in the comments.