The RTO/RPO conversation with the business
Business stakeholders have opinions about tolerance for downtime and data loss. Those opinions are usually "zero." Your job is to get to realistic numbers.
The wishful phase
Left to their own devices, business stakeholders will answer:
- Downtime tolerance: 0 minutes
- Data loss tolerance: 0 seconds
That's not actionable. It's also usually not what they actually want once they understand the cost.
The pricing phase
Show them this table:
| RTO | RPO | Strategy | Relative annual cost |
|---|---|---|---|
| 24+ hrs | 24 hrs | Backup & Restore | 1x baseline |
| 4 hrs | 1 hr | Pilot Light | 2x baseline |
| 15 min | 5 min | Warm Standby | 4x baseline |
| < 1 min | < 1 min | Active/Active | 8x baseline |
Now the conversation changes. "You're saying it would cost 4x to go from 4-hour to 15-minute RTO? Let's talk about which systems really need that."
The per-system conversation
Not every system has the same RTO/RPO. A customer-facing payment system has dramatically different needs than an HR vacation tracker. Build a table:
| System | Business impact | RTO target | RPO target | Strategy tier |
|---|---|---|---|---|
| Payments | Revenue loss + PR | 5 min | 1 min | Warm standby |
| Customer portal | CX impact | 30 min | 5 min | Pilot light |
| Internal ticketing | Productivity | 24 hr | 24 hr | Backup & Restore |
| Analytics DW | Delayed reports | 72 hr | 24 hr | Backup & Restore |
The test commitment
Once RTO/RPO are set, commit to testing quarterly. Without tests, the numbers are fiction.