Cloud should compound value, not cost. In South Africa, the fastest way to prove cloud’s worth is to cut waste without slowing delivery. Here’s the practical 30-day playbook we run with clients.
Where the Spend Leaks
- Idle dev/test: 9-to-5 teams, 24×7 environments.
- Over-sized instances: CPU avg < 20%, memory < 40%.
- Forgotten storage: unattached volumes, old snapshots, multi-AZ by accident.
- Chatty networks: inter-AZ/egress fees due to poor placement.
- Always-on analytics: clusters up for convenience.
- Lift-and-shift ghosts: legacy licenses & VM sprawl.
- Discounts left on table: no Savings Plans/Committed Use.
30-Day Cost Playbook (No Slowdown)
- Days 1–3: turn on detailed cost & usage; tag % coverage baseline; freeze new instance families.
- Days 4–10: right-size top 20% by cost; schedule dev/test off (night+weekends); delete orphaned storage.
- Days 11–20: placement groups to reduce cross-AZ; autoscaling policies; pause idle analytics.
- Days 21–30: buy 1-year commitments for steady 60–70% baseline; publish a “run-rate” graph weekly.
Target: 15–30% run-rate reduction in 30 days, zero feature delay.
Cost Scoreboard
- Run-rate (R/mo): last 4 weeks vs baseline
- Idle %: hours off / hours total in non-prod
- Rightsized %: instances within ±1 size of target
- Discount coverage: committed/covered usage %
Want a 30-day savings sprint?
We’ll baseline, execute, and report weekly — without slowing delivery.
Start a cost sprint