Complete cost allocation across 156 engineers and 20 product lines made higher-order AI and unit economic efficiency goals possible
The Challenge
Founded in 1981, Progress Software enjoys the rare distinction of successfully transitioning from a deployed model to a SaaS model. Powered by acquisition-based growth, Progress had 20 distinct product lines by late 2024 and a fair amount of cloud cost chaos.
“The central premise as related to cloud costs was, ‘Don’t get in the business’s way,’” said Greg Colletti, Progress’ director of strategic portfolio management, and the owner of Progress’ cloud cost management (CCM) project. “Our engineers were building cool stuff, but we lacked a way to show them that every engineering decision — especially with AI now — is a monetary decision.”
To manage their cloud costs — which spanned AWS, GCP, Azure, OCI, and IBM — Progress provisioned a legacy CCM platform and had also engaged a managed service provider (MSP) to handle their reserved instances (RIs) and savings plans (SPs). But this approach ultimately proved ineffective. The RI and SP discount driven by the MSP was offset by the MSP’s fee, and the legacy CCM platform’s innovation slowed dramatically after getting acquired by a large software conglomerate.
“The straw that broke the camel’s back was when we tried to mimic the structure of our business in allocation,” Greg said. “With our former CCM, this was incredibly laborious, inflexible, and never quite complete. It took the better part of a quarter to get that structure in place, and then it was almost impossible to update amid natural shifts in our business.”
In a new CCM partner, Progress was looking for several key things:
- Allocation: Progress knew the allocation framework for organizing their cloud costs, but needed a partner that could bring this framework to life.
- Engineering engagement: Progress wanted to shift from cost-agnostic engineering to cost-conscious engineering.
- AI cost efficiency: Progress had named a Chief AI Officer shortly before beginning the CCM search. Their AI spending was about to ramp up significantly, and as it did, they wanted to lay the groundwork for long-term efficiency.
- Unit economics: Ultimately, Progress wanted to use their new CCM platform to give them a clear line of sight into absolute business efficiency metrics, including margin per product and margin per feature.
The Solution
Progress identified more than a dozen potential CCM vendors, and then narrowed the longlist to three, including CloudZero. To evaluate these three vendors, Greg engaged engineering leaders from 12 of the 20 Progress product lines, and supplied them with a 39-point evaluation schema based on their most urgent business needs. After a two-week process that could well be considered a CCM boot camp, CloudZero emerged as the democratically elected favorite.
“The deciding factor was really automated cost allocation,” Greg said. “The other two platforms struggled to automate it. If we want engineers to care, we have to give them easy access to reliably accurate cost data. CloudZero was the decisive favorite for fast, accurate visibility.”
The Results
At the time of writing, Progress has been a CloudZero customer for just under a year. Their main achievements break into four categories:
- Allocation: Automated organization-wide cloud cost visibility and surface cost savings opportunities
- Engineering Engagement: Reached 156 daily active users in CloudZero
- Unit Economics: Established four unit cost metrics for key products
- AI Cost Efficiency: Prevented a Claude service from accruing unnecessary costs
Allocation: Automated org-wide cloud cost visibility
Progress uses CloudZero to automatically allocate cloud costs based on an internal hierarchy that wasn’t possible to model with their former CCM vendor. This includes such categories as:
- Cost center
- Product code
- Business owner
- Business unit
- Product line
- Environment
Fast, flexible, accurate allocation has made it possible for Progress to give engineers constant access to the costs that matter most to them.
This has also resulted in a more proactive approach to cost savings.
“We’ve caught a lot of cost anomalies with CloudZero,” Greg said, also explaining that now, engineers find and raise savings opportunities proactively. “Just yesterday, someone flagged a $40k spike in cloud spend, and we were able to shut down the underlying source to prevent any additional cost.”
Engineering Engagement: Reached 156 daily active CloudZero users
Because Progress allocated their costs by business unit, product line, and owner, it was easy for them to directly engage the engineers responsible for cloud costs within each of those dimensions.
“We’re building FinOps champions,” Greg said. “People are taking advantage of CloudZero Academy. They’re asking the right questions. They’re proactively monitoring their costs. We only expect the number of daily active users to keep increasing.”
Unit Economics: Established four unit cost metrics for two key products
Progress has already put a major dent in their unit economics goals, establishing the following four unit cost metrics for Progress® Data Cloud® and Progress® ShareFile®, two of their flagship products:
- Cost per customer (per product)
- Cost per network ingest (per product)
- Cost per click (per product)
- Cost per unit of storage (per product)
“Cloud unit economics will help us monitor customer usage habits, build some controls into our platform, and ultimately, refine our pricing model. Right now, it’s about showback and awareness. Next, it’ll get to pricing and packaging,” Greg said.
AI Cost Efficiency: Prevented a Claude service from accruing unnecessary costs
CloudZero has already helped Progress keep AI costs under control. Recently, AI surfaced a cost anomaly for a Claude service in Progress’ testing environment. Its costs were mounting on an exponential curve, but the CloudZero anomaly enabled them to cut it off.
“CloudZero enables us to tell whether we’re using AI tools responsibly,” Greg said.
CloudZero gave Progress the structure and transparency needed to make cloud costs a day-to-day part of engineering decision-making. With allocation mapped to how the business actually operates, teams can spot anomalies faster, uncover savings opportunities and measure unit costs with confidence. That foundation is helping Progress pursue AI-forward growth with greater control and clarity.
