Why Faster CPUs Can Be Cheaper: Real World FinOps Lessons ft. Matan Ben-Ishay, Credit Karma | Ep #70
What if your biggest cloud inefficiency isn’t how much you’re using, but how long your workloads are running?
In this special episode of FinOps in Action, PointFive’s Co-Founder and CPO, Gal Ben David sits down with Matan Ben-Ishay, leading FP&A, Engineering, and Cloud Infrastructure at Credit Karma, to unpack a real-world cloud efficiency discovery that led to tens of millions of dollars in savings without reducing usage or cutting performance.
Matan shares how a simple assumption about “cheaper” CPUs quietly inflated costs, why billing data alone didn’t reveal the problem, and how the true FinOps impact only happens when finance and engineering work side by side.
We also covered how FinOps wins happen at the intersection of finance, engineering, and product and how understanding accounting concepts is a career advantage in FinOps.
Here’s what we talked about:
How a hidden GCP dataflow inefficiency cost tens of millions until Matan sat with engineers and discovered faster CPUs actually cost less per hour
Why rigid governance policies backfire and how flexible guidelines that encourage engineers to evaluate their actual workload needs prevent waste
Why FinOps professionals need to understand accounting concepts (depreciation, amortization, commitments)
The emerging need for cloud efficiency engineers embedded in teams who can bridge the gap between FinOps strategy and hands-on optimization work
How running parallel tests of old vs. new approaches proves value to skeptical engineers and turns short-term testing costs into long-term savings
The structural friction between engineering OKRs (focused on launches and reliability) and cost efficiency, and why engineers should treat cloud spend like any other company resource
“We need those cloud efficiency engineers in order to understand.” - Matan Ben-Ishay
Connect with Matan:
Website: https://www.creditkarma.com/
In this episode of FinOps in Action, Matan Ben-Ishay draws on experience spanning EY cloud consulting to FP&A engineering at Credit Karma to argue that misaligned incentives, not technical complexity, are FinOps practitioners’ biggest obstacle. He breaks down why giving engineers public credit for cost optimization is the key to driving org wide adoption, how three year commitment forecasting with capacity planning and finance creates a true 360 degree view of infrastructure decisions, and why governance best practices must be embedded from day one, because the engineer who skips retention policies today becomes the manager who doesn’t care about them tomorrow.