A founder told me yesterday:
"We just bought a bigger Savings Plan. Our AWS bill should drop next month, right?"
I told him: "Probably not."
And he went quiet.
Because every FinOps vendor is saying:
"Buy commitments to save"
"Savings Plans = instant 30% off"
"Lock in, watch costs fall"
But here's the truth nobody explains:
A Savings Plan doesn't lower your bill. It lowers the rate on the compute you actually run inside the commit window. That's a very different thing.
If your coverage drifts → the discount evaporates
If Karpenter changes instance mix → your commit stops matching your shape
If Finance buys without telling engineering → you pay for capacity nobody uses
60% of Indian SaaS teams we audit have Savings Plan coverage under 50%, even when they just "bought more."
The issue isn't the commit. The issue is the drift.
Real cost control starts here:
→ Monitor coverage % weekly, not quarterly
→ Tag every capacity change that can shift it
→ Right-size the commit to actual p50 shape, not aspirational peak
→ Then automate the renewal
That's when Savings Plans actually print ROI. Not before.
Most teams don't need a bigger commit. They need visibility into the drift.
If this made you rethink that ₹40L commit in your console…
Repost. There's a CFO in your network about to renew a Savings Plan that was never working.
Top comments (0)