Every recovery plan starts with two questions. The answers have names, and they pull in opposite directions.
RPO defines the last acceptable recovery point. RTO defines the maximum acceptable recovery duration.
The same two questions, answered for a business where the stakes are low.
Picture a one-chair barber shop. A wall of regulars, and a tablet by the mirror that holds the bookings and takes the card payments. Every night, when the shop locks up, the tablet copies the day to the cloud.
One afternoon at two, the tablet dies.
Two questions decide how bad this really is, and for this shop, both answers are forgiving.
Only the morning's bookings, taken since the shop opened, are gone — about four hours the owner can rebuild from memory and a stack of text messages. Losing that much is manageable, so the data target stays loose.
For the rest of the afternoon the shop runs on a paper notebook, and the tablet is working again by early evening. A few hours of disruption stings, but the shop absorbs it, so the downtime target stays loose too.
Small stakes, loose numbers. Put the same two questions to a hospital or a bank, and both shrink toward zero.
Design a recovery, then run it, and see whether the recovery target fits the business.
The situation. A failure has just taken a business offline. Choose how much data it can lose and how long it can stay down, then run the recovery. Reach a protected outcome. You have two attempts.