Post-Acquisition Cloud Optimization: The First 90 Days

Post-Acquisition Cloud Optimization: The First 90 Days

The first 90 days after acquisition are critical for establishing momentum on cloud cost optimization. Done well, a structured 90-day program can deliver 20-35% reduction in cloud spending -- generating immediate EBITDA improvement while building the foundation for sustained financial discipline.

I have led this playbook at dozens of PE-backed companies, and the pattern is remarkably consistent. The opportunities are predictable, the implementation is straightforward, and the results are measurable within weeks. Here is the phase-by-phase approach.

Before Day 1: Pre-Close Preparation

Effective cloud optimization starts before the deal closes. During diligence, you should have already gathered:

• 12 months of cloud billing data (monthly granularity at minimum)
• The current AWS account structure and organization
• A high-level architecture overview
• Information about existing Reserved Instances, Savings Plans, or EDP commitments
• Identification of the key technical contacts who manage cloud infrastructure

With this information, you can build a preliminary optimization model that quantifies the expected savings by category and sets realistic targets for each phase. This model becomes the scorecard against which you track progress.

Phase 1: Days 1-30 -- Visibility and Quick Wins

Objective: Establish full visibility into cloud spending and capture the easiest savings opportunities.

Expected savings: 10-15% of total cloud spend.

Week 1: Discovery and Instrumentation

Your first priority is gaining access to the cloud environment and understanding what you are working with. Key actions:

• Enable AWS Cost Explorer and Cost and Usage Reports (CUR). If the company does not have these configured, turn them on immediately. CUR provides the granular billing data needed for detailed analysis.
• Inventory all AWS accounts and resources. Use AWS Resource Groups Tag Editor or a third-party tool to enumerate every running resource across all accounts.
• Assess current tagging coverage. Determine what percentage of resources have cost allocation tags and what tag keys are in use.
• Identify the top 10 spending categories. Sort the AWS bill by service and identify the largest cost drivers. In most environments, EC2, RDS, S3, and data transfer account for 70-80% of total spend.

Weeks 2-3: Zombie Resource Elimination

With inventory in hand, systematically identify and terminate unused resources:

• Idle EC2 instances: Query CloudWatch for instances with less than 2% average CPU utilization over the past 14 days and zero network traffic. Validate with the engineering team, then terminate. Expected savings: $2,000-$20,000/month depending on environment size.
• Unattached EBS volumes: These are storage volumes not connected to any running instance. They accumulate over time as instances are terminated but their volumes are not. Expected savings: $500-$5,000/month.
• Old EBS snapshots: Identify snapshots older than 90 days with no lifecycle policy. Review and delete unnecessary snapshots. Expected savings: $500-$8,000/month.
• Idle Elastic Load Balancers: Load balancers with zero healthy targets or zero request count. Expected savings: $200-$2,000/month.
• Unused Elastic IPs: AWS charges $3.60/month for each allocated but unattached Elastic IP. Small per-unit cost but they accumulate.

Week 4: Storage Optimization

• Implement S3 lifecycle policies. Move objects to Infrequent Access (IA) tier after 30-60 days and to Glacier after 90-180 days, depending on access patterns. For companies with large S3 footprints, this alone can save 40-60% on storage costs.
• Review EBS volume types. Many companies use Provisioned IOPS (io1/io2) volumes where General Purpose (gp3) would be sufficient. GP3 offers 3,000 IOPS and 125 MB/s throughput included in the base price -- sufficient for most workloads -- at a significant cost reduction.
• Optimize S3 request costs. For high-request-volume buckets, evaluate whether S3 Express One Zone or S3 Intelligent-Tiering would be more cost-effective.

Phase 1 deliverables: Complete resource inventory, tagging assessment report, zombie resource elimination, storage optimization. Monthly run-rate savings of 10-15%.

Phase 2: Days 31-60 -- Right-Sizing and Commitments

Objective: Optimize running resources and implement commitment-based pricing.

Expected savings: Additional 10-15% of total cloud spend (cumulative 20-30%).

Weeks 5-6: Instance Right-Sizing

With 30 days of CloudWatch data since your initial instrumentation, you now have solid utilization baselines for every running instance.

• Identify right-sizing candidates. Flag instances where peak CPU utilization over 30 days is below 40% and average is below 20%. These are candidates for downsizing -- typically by one or two instance sizes.
• Evaluate instance generation. Many companies are running previous-generation instances (c5 instead of c7g, m5 instead of m7g). Current-generation instances offer 20-40% better price-performance. Migration is typically low-risk for stateless workloads.
• Consider Graviton (ARM) instances. AWS Graviton-based instances (c7g, m7g, r7g) offer approximately 20% cost savings over comparable x86 instances with equal or better performance. If the application stack supports ARM (most Linux workloads do), this is a reliable cost reduction with minimal risk.
• Right-size RDS instances. Databases are frequently the most over-provisioned resources. Analyze CPU, memory, and IOPS utilization. Downsize where safe, and consider Aurora Serverless for variable-demand database workloads.

Weeks 7-8: Commitment-Based Pricing

With right-sizing complete (so you are not committing to resources that will change), implement commitment-based pricing:

• Compute Savings Plans cover EC2, Fargate, and Lambda usage with 1-year or 3-year terms. They offer flexibility across instance families and regions. Start here for most workloads.
• EC2 Instance Savings Plans offer deeper discounts but are locked to a specific instance family in a specific region. Use for workloads where you have high confidence in long-term stability.
• Reserved Instances for RDS and ElastiCache, where Savings Plans do not apply.

Target coverage: 60-70% of stable compute spend should be covered by commitments. Leave 30-40% as On-Demand to maintain flexibility for variable and growing workloads. Over-committing is a real risk -- if usage drops below commitment levels, you pay for unused commitments.

Phase 2 deliverables: Right-sizing completion report, commitment-based pricing implementation, updated savings forecast. Monthly run-rate savings of 20-30%.

Phase 3: Days 61-90 -- Architecture and Governance

Objective: Implement structural improvements and establish ongoing governance.

Expected savings: Additional 5-10% of total cloud spend (cumulative 25-35%).

Weeks 9-10: Autoscaling and Scheduling

• Implement autoscaling for variable-demand workloads. Configure target-tracking scaling policies based on CPU utilization, request count, or custom metrics. Start conservatively (scale out aggressively, scale in gradually) and tune over time.
• Schedule non-production environments. Implement automated start/stop schedules for development, staging, and QA environments. Run them 10-12 hours on weekdays only (unless overnight testing is required). Expected reduction: 65-70% of non-production compute costs.
• Implement spot instances for fault-tolerant workloads like batch processing, CI/CD, and development environments. Spot pricing is typically 60-90% below On-Demand.

Weeks 11-12: FinOps Governance Framework

The final -- and arguably most important -- phase is establishing the governance framework that will sustain your optimization gains.

• Implement mandatory tagging policies. Use AWS Config rules or Service Control Policies to prevent the creation of resources without required tags.
• Set up automated cost anomaly detection. AWS Cost Anomaly Detection or third-party tools should alert on unexpected spending spikes within hours, not at month-end.
• Establish a FinOps cadence: Weekly cost reviews with engineering leads, monthly reviews with finance, quarterly executive reviews.
• Define cloud budgets by team and environment. Track actual vs. budget monthly and require variance explanations.
• Create a FinOps dashboard accessible to all stakeholders -- engineering, finance, and executive leadership.

Phase 3 deliverables: Autoscaling and scheduling implementation, FinOps governance framework, ongoing optimization playbook, final savings report.

Expected Savings Timeline

Based on my experience across dozens of engagements, here is what a realistic savings timeline looks like for a company spending $200K/month on AWS:

These are conservative estimates for a company that has never undergone systematic optimization. Companies with larger environments or more significant waste can see proportionally larger returns.

Sustaining the Gains

The 90-day program is a sprint, but cloud optimization is a marathon. Without ongoing governance, savings erode within 6-12 months as new resources are provisioned without oversight, commitments expire without renewal, and the organization gradually returns to its old habits.

The governance framework established in Phase 3 is your insurance policy. Invest in it, maintain it, and hold the management team accountable for cloud cost performance just as you would any other financial metric.

Ready to evaluate cloud economics in your next deal? Book a free discovery call to discuss your specific situation.