Platform Reliability Score Calculator

Calculate an overall platform reliability score from availability, MTTR, incident rate and error budget consumption, mapped to a letter grade.

Inputs

Availability%

Measured uptime/availability over the period.

MTTRhours

Mean Time to Recovery/Resolution for incidents.

Incidents per Monthincidents

Average number of production incidents per month.

Error Budget Consumed%

Percent of the period's error budget already consumed.

Platform Reliability Score

89.4

Reliability Grade

Availability Sub-Score

99.9

MTTR Sub-Score

93.8

Incident-Rate Sub-Score

85.0

Error-Budget Sub-Score

60.0

Step by step

Availability score (40% weight)
99.9% × 0.40
= 39.96
MTTR score (25% weight)
max(0, 100 − 1.5/24×100) × 0.25
= 23.44
Incident-rate score (20% weight)
max(0, 100 − 3/20×100) × 0.20
= 17.00
Error-budget score (15% weight)
max(0, 100 − 40) × 0.15
= 9.00
Overall reliability score
sum of weighted sub-scores
= 89.4 (B)

How it works

This composite score blends four reliability signals into a single weighted number: availability (40% weight, since uptime is the most visible reliability signal), MTTR (25%, rewarding fast recovery), incident rate (20%, penalizing frequent instability), and error budget consumption (15%, reflecting SLO health). Each sub-score is normalized to 0-100 before weighting, and the final score maps to a letter grade (A: 95+, B: 85+, C: 70+, D: 50+, F: below 50).

Formulas

MTTR sub-score

mttr_score = max(0, 100 - (mttr_hours / 24) × 100)

mttr_hours: Mean Time to Recovery in hours

Overall reliability score

score = availability × 0.4 + mttr_score × 0.25 + incident_rate_score × 0.2 + budget_score × 0.15

availability: Measured availability percentage (0-100)
mttr_score: MTTR sub-score (0-100)
incident_rate_score: max(0, 100 - incidents_per_month / 20 × 100)
budget_score: max(0, 100 - error_budget_consumed_percent)

Frequently Asked Questions

Why is availability weighted highest?

Availability is the most direct, universally understood measure of whether users can access the service, so it carries the largest weight (40%) while the other signals capture how well incidents are handled and how much reliability margin remains.

How are the MTTR and incident-rate sub-scores normalized?

MTTR score linearly decays from 100 (instant recovery) to 0 at a 24-hour resolution time; incident-rate score linearly decays from 100 (zero incidents) to 0 at 20 incidents/month — both are heuristic reference points and can be recalibrated to your organization's norms.

Can this replace SLO reporting?

No — this is a simplified composite scorecard useful for at-a-glance comparisons across teams or time periods. Formal SLO compliance should still be tracked per-service against its specific SLI/SLO definitions.

What should a team do with a 'D' or 'F' grade?

Treat it as a signal to prioritize reliability work: identify which sub-score is dragging the total down (e.g. high incident rate vs. slow MTTR) and target that specific area — a low incident-rate score points to stability work, while a low MTTR score points to on-call/runbook improvements.

You might also need

SLI Calculator Change Failure Rate Calculator Cluster Health Score Calculator