Calculating ROI for an SRE Agent Project

Four cost lines, three benefit lines, and the assumption that ruins the math if you get it wrong. The calculator, with defaults, that gets you to a defensible number.

Benefits

MTTR reduction: minutes saved per incident × incidents per month × cost-per-minute-of-downtime.

On-call burden reduction: hours saved per week × hourly cost × engineers.

Postmortem speed: hours saved per postmortem × postmortems per month.

Costs

Engineering build/maintenance: FTE-equivalent for the team.

Vendor or compute spend: model API calls, infrastructure.

Onboarding cost: training the rest of the team to work with the agent.

Risk cost: occasional agent errors that require remediation.

The assumption that ruins the math

Most ROI calculators assume the agent handles every relevant incident. They do not. Apply a coverage multiplier.

Year 1 coverage is realistically 30-50% of in-scope incidents. The other 50-70% still need humans.

Year 3 coverage approaches 70-90% with mature workflows. Don't model year-1 numbers as steady-state.

The calculator

ROI = (annual benefit × coverage) − annual cost.

Default inputs: 3 engineers building, $300k each fully-loaded; 100 incidents/month relevant; 30 minutes saved per handled incident; $500/min downtime cost.

These numbers cluster around "break-even in 18 months" for most teams. Sensitivity: if downtime cost is much higher (regulated industries), break-even is 6-9 months.

Be conservative on day one

Don't promise 80% MTTR reduction in year one. The team won't trust it; the data won't justify it.

Promise 30% in year one, 50% in year two. Underpromise and overdeliver; the budget approval is more reliable when the math is conservative.

Track actual numbers monthly. Adjust the public ROI claim as data arrives.