Service Level Agreement
TL;DR: heft.io publishes per-service monthly uptime targets from 99.5% to 99.99%. Planned maintenance is announced in advance. Live status is at status.heft.io.
What's covered
This SLA covers uptime for:
- Tracking JS
Client-side tracker script attrack.heft.io/lib.js - Tracking API
Event collection attrack.heft.io - Data API
Analytics queries and dashboard data atapp.heft.io - Authentication
Sign-in and session management atapp.heft.io - Site sharing
Public analytics pages atsite.heft.io - Static application and website
Static assets and pages atapp.heft.ioandheft.io
Documentation pages and email delivery are outside this SLA. Beta features, preview releases, and anything labelled experimental are excluded until they reach general availability.
Uptime targets
heft.io commits to the following monthly availability rates. The measurement period is each calendar month, UTC. On a 30-day month, the targets allow at most roughly:
| Service | Monthly availability | Max. unavailability (30-day month) |
|---|---|---|
| Tracking JS | 99.99% | ~4 minutes |
| Static application and website | 99.99% | ~4 minutes |
| Tracking API | 99.9% | ~43 minutes |
| Data API | 99.5% | ~3.6 hours |
| Authentication | 99.5% | ~3.6 hours |
| Site sharing | 99.5% | ~3.6 hours |
Scheduled maintenance windows described below are excluded from these calculations.
How availability is measured
- Unavailability
- The covered service can't be reached, or returns server errors (
5xx) for valid requests, continuously for at least five minutes. Monitoring has to confirm the incident. - Valid request
- A request that would normally succeed when the service is healthy – for example, a
GETfortrack.heft.io/lib.js, sign-in with correct credentials, a tracking request with a valid site identifier, or an analytics query with a valid session.
Monthly availability rate
- Ttotal
- Total minutes in the calendar month.
- Tdown
- Minutes of confirmed unavailability in that month.
Maintenance
Planned maintenance covers security patches, infrastructure updates, and database operations. Where possible, it's scheduled during the lowest-traffic windows derived from aggregate traffic patterns across sites tracked with heft.io. Maintenance is announced at least 48 hours in advance on status.heft.io. Account holders get an email too.
Emergency maintenance for a security or stability issue may run without prior notice. Downtime during emergency maintenance still counts towards the monthly availability rate.
Exclusions
The uptime targets don't apply when downtime is caused by, or related to:
- Factors outside heft.io's reasonable control: force majeure, upstream internet or DNS failures outside heft.io's network, or DDoS attacks mitigated at the edge
- Your configuration, integration, or misuse of the service: invalid API keys, rate-limit breaches, or embedding errors on your site
- Suspension or termination under the Terms of Service
- Scheduled maintenance announced in accordance with this SLA
Status and monitoring
Service health is monitored continuously. Subscribe at status.heft.io for incident history and maintenance notices.
Changes
This SLA may be updated. If something material changes, you'll hear about it in advance. Continued use means you accept the latest version.
Contact
Questions about this SLA? Use the contact options on the site.
Last updated: May 2026