What Cloud Exit Calculates
The month your cumulative cloud bill overtakes self-hosted cost including depreciation, electricity, and risk.
Cloud Exit answers one question: when does leaving the cloud actually pay off?
The cloud bill grows quietly until someone asks if a $2k/month AWS line item could be replaced with an $8k rack. The honest answer requires loading the self-host side with electricity, maintenance labor, and uptime risk — and loading the cloud side with egress fees. Cloud Exit does both.
Quick answer
Cloud Exit answers one question: when does leaving the cloud actually pay off?
Key points
- ▸ Formula (cloud): (Monthly + Egress) × Months.
- ▸ Formula (self-host): (Hardware ÷ Depreciation Months) × Months + Electricity × Months + Maint Hours × Hourly Rate × Months + Uptime Risk % × Cloud Monthly × Months.
- ▸ Egress is the silent killer. Data transfer out of the big three clouds is $0.08-0.12/GB — often 20-40% of the total cloud bill at scale.
- ▸ Hardware depreciation defaults to 36 months. Use 24 for bleeding-edge, 48-60 for low-duty.
- ▸ Uptime risk captures the "insurance premium" you pay cloud for five-nines reliability. Model it as a % of cloud spend you're effectively saving for self-host outages.
Examples
- $2k cloud + $250 egress vs $8k rack, 36-mo depreciation, $100/mo elec, 4hr/mo maint @ $80/hr, 10% uptime riskCloud $2,250/mo. Self-host $222 + $100 + $320 + $200 = $842/mo. Crossover at month 4-5 — hardware pays itself off fast once egress is in the mix.
- Low-egress workload: $1k cloud + $20 egress vs $5k rackCloud $1,020/mo. Self-host ~$700/mo. Crossover at month 8-10 — less dramatic but still favors exit over the hardware lifetime.
- High-compute, low-transfer: $3k cloud + $40 egress vs $15k workstationCloud $3,040/mo. Self-host ~$950/mo. Crossover at month 7. Self-hosting wins big — but only if the maintenance hours and downtime are acceptable.
When to use which tool
- CYAN · STABLE — Crossover under 6 months — exit now, hardware pays back within two quarters.
- GOLD · GUARDED — Crossover 6-18 months — viable for steady workloads, risky if traffic is spiky.
- MAGENTA · CRITICAL — Crossover above 24 months or never — stay in cloud, self-host penalty too steep.
Related
- The Cloud ExitWhen does a $2k cloud bill justify an $8k rack? Crossover math with depreciation, electricity, and uptime risk.
- When to Run Cloud ExitFive specific moments when the cloud-vs-self-host math shifts enough to force the question.
- Six Cloud Exit MistakesThe errors that make self-hosting look cheaper than it turns out to be — and one that makes cloud look cheaper than it is.
Frequently asked questions
› Does the cloud still win for spiky workloads?
Yes. Elasticity is the cloud's real value — workloads that 10× peak don't justify self-hosting even when the monthly baseline suggests otherwise. Cloud Exit is for steady-state workloads.
› What about hybrid setups?
Model the cloud portion as cloud-only, the steady workload as self-host, and sum. Hybrid is often the right answer for businesses with mixed workload shapes.
› Does this account for engineer time to operate the self-hosted system?
Yes, via Maint Hours × Hourly Rate. Use loaded engineer cost, not base salary. Underestimating maintenance hours is the most common Cloud Exit error.
› How should I use a decision framework in real life? How-to
Use a decision framework to expose the tradeoff, not to outsource the decision. Write down the inputs, compare the output with your constraints, then ask what would change the answer. The strongest use is scenario testing: base case, conservative case, and failure case.
› Is this financial, legal, or tax advice? Trust & accuracy
No, this is not legal, financial, tax, medical, or professional advice unless the page explicitly says that use case is supported. It organizes assumptions so you can inspect them. Verify high-stakes choices with qualified people who can review facts, contracts, regulations, and downside risk.