Overview
Zebra Technologies designs hardware, software, automation solutions, and services for the frontline, spanning retail, healthcare, manufacturing, and supply chain and inventory logistics. To support large-scale operations that demand performance, reliability, and security, Zebra runs a hybrid estate that includes AWS, Azure, and Google Cloud alongside on-prem environments. Optimizing costs requires a FinOps approach that can keep pace with dynamic workloads.
Dean Pereyra, Manager of Cloud Enablement and Engineering at Zebra Technologies, leads a centralized team and helps Zebra adopt best-in-class cloud services to drive operational excellence and cost optimization. Beyond FinOps, his team also supports architecture, security, governance, and AI workloads.
With many priorities competing for time, Zebra relied on automated solutions to reduce repetitive work and keep optimization on track. Zebra’s team initially used ProsperOps to automate rate optimization in AWS, while a dedicated internal resource managed the ongoing commitment work across their broader cloud footprint.
Challenges with Resource Risk and Kubernetes Optimization
Zebra’s FinOps team faced operational risk when their resource for rate optimization left the company. Without a dedicated owner, it became challenging to keep up with commitment renewals in Azure and Google Cloud. Scaling rate optimization across a multi-cloud footprint was already difficult with limited bandwidth. Each cloud provider also had different discount instruments (e.g., Azure Reservations, AWS Reserved Instances, Google Cloud resource-based Committed Use Discounts), pricing structures, and rules.
The team was also rightsizing Kubernetes workloads on Google Cloud, which meant they expected usage to decrease. Zebra needed to keep rate optimization synchronized with those workload changes. If Zebra purchased commitments based on yesterday’s usage while optimizing clusters, they risked underutilized commitments and wasted spend.
Zebra followed the best practice of centralized rate optimization to maximize savings, but engineering changes to usage were decentralized. Keeping the commitment strategy in sync with usage required frequent check-ins with each business unit’s engineering team. These meetings consumed a significant amount of time, and if usage changed faster than the corresponding changes for commitments, there was risk of overcommitment.
Having a sophisticated FinOps solution that could adapt to ongoing environment changes would help business units stay within budget.
Expanding ProsperOps from AWS to Multi-Cloud Rate Optimization
Dean needed a solution to manage commitments across Azure and Google Cloud without adding headcount. Because ProsperOps was already delivering strong AWS savings, Zebra expanded ProsperOps to cover Azure and Google Cloud.
ProsperOps continuously monitored usage and automatically adjusted Zebra’s commitments as conditions changed in Zebra’s multi-cloud environments. This also helped reduce meetings with the engineering and finance teams of each business unit.
Zebra achieved even greater compute savings through ProsperOps. ProsperOps’ sophisticated algorithms maximized flexibility by carefully structuring commitments, so that they could be easily modified.
For example, ProsperOps built adaptive Reservation ladders in Zebra’s Azure compute environment by deploying small increments of commitment over time to create multiple expiration points, rather than infrequent batch purchases. On Google Cloud, ProsperOps helped Zebra modify resource-based Committed Use Discounts (CUDs), allowing Zebra to receive maximized compute discounts without taking on full 1-year/3-year terms. ProsperOps also utilized Convertible Reserved Instances (CRIs) to easily adapt to fluctuations in AWS compute, which both maximized Effective Savings Rate (ESR) and decreased commitment risk.
Zebra improved savings and increased coverage with ProsperOps (in green) for their Azure compute environment.
Using ProsperOps, Zebra maintained high coverage for their Google Cloud compute environment after their resource left.
Zebra also significantly improved their commitment management on AWS using ProsperOps.
Higher ESR Across AWS and Azure, Strong Performance Maintained on Google Cloud
Zebra improved its all-discounts compute Effective Savings Rates (ESRs) across AWS and Azure and maintained strong outcomes on Google Cloud. The all-discounts compute ESRs shown below included savings from commitments, negotiated rates, and spot VMs. Coverage on AWS and Azure also improved.
| Zebra’s Cloud Environments | Compute ESR Before | Compute ESR After | Difference |
| Azure | 22.5% | 31.2% | +8.7% |
| AWS | 16.7% | 29.9% | +13.2% |
| Google Cloud | 50% | 51% | +1% |
Summary
Zebra consistently achieved greater savings with ProsperOps, which allowed Dean and his team to plan and forecast more confidently for each business unit. Automation also reduced the back-and-forth meetings with engineering and finance teams.
“I looked for a solution that could scale across all three hyperscalers. ProsperOps met this need. It continuously monitored usage for all of Zebra’s business units and adjusted commitments as conditions changed, without requiring manual effort.”
This freed-up time is reinvested into high-impact initiatives where human judgment matters most, including strengthening governance frameworks and policies as well as migrating business units to new tools.
Start a free cloud savings analysis with ProsperOps to see how much additional savings and flexibility you could unlock across AWS, Azure, and Google Cloud, without adding overhead.
