All posts
Case Study From the archive November 4, 2024 · 6 min read

Case study: how a cybersecurity company found $3M in cloud savings

A cybersecurity company spending $15M a year on cloud infrastructure had no per-product visibility and no commitment strategy. Three months later, they had both — and roughly $3M in annual savings.

Tm
The moneta Team
moneta team

The starting point: a $15M black box

The customer was a mid-market cybersecurity company running across multiple product lines, dozens of microservices, and three cloud regions. Annual cloud spend: roughly $15 million and rising.

Two problems compounded each other. First, costs weren't attributed to products — the executive team could see the total, but not which products were profitable and which were subsidized. Second, the commitment portfolio was minimal; most workloads ran on on-demand pricing even though usage patterns were predictable.

Step one: tag everything that moves

Before any optimization work, the team built a tagging strategy. Every resource got a product, team, and environment tag at provisioning time. Untagged resources were flagged and assigned an owner within two weeks. By the end of month one, unallocated spend had dropped from roughly 40% of the total to under 5%.

This alone changed the conversation. Product managers could see what their work cost. Engineering leads could see where their team's spend was going. Finance could finally produce a per-product P&L that included cloud as a real input.

Step two: build the commitment portfolio

With usage patterns finally visible, the team modeled which workloads were steady enough to justify commitments and which should stay on-demand. Reserved Instance coverage went from roughly 20% to over 70%. A Savings Plan layered on top covered the rest of the compute footprint.

The math worked out to about $3 million in annualized savings — without any reduction in capacity. The same workloads ran on the same instances; they just cost less because the commitment portfolio was now matched to actual consumption.

Step three: forecasting and accountability

The last step was making cloud spend a forward-looking number, not a backward-looking one. The team set monthly budgets per product, tied them to a forecasting model that used historical usage and known roadmap events, and published variance against budget every Monday.

Forecast accuracy improved from a roughly 25% miss to under 8% within two quarters. More importantly, engineering teams started having proactive conversations about cost: "we're trending 12% over for the quarter, here's what we can defer." Those conversations had been impossible when nobody knew what anything cost.

What the executive team got

The CFO got three things: complete visibility into where $15 million was going, $3 million in annualized savings, and a financial reporting cadence that matched the rest of the business. None of it required a re-platform, a re-org, or a multi-year initiative. It required a discipline, applied consistently for ninety days.

If your cloud spend looks like a black box from the executive view, the path forward is similar: tag, commit, forecast. In that order.

Want this in practice?

See moneta in 30 minutes.

Walk through how moneta connects billing, pricing, discounts, and customer-level margin — with your numbers, not ours.

Book a demo