Managing Technical Debt: A Practical Guide for Product Teams
Learn how to spot, measure and pay down technical debt without freezing your roadmap, and how to manage it when part of your engineering is outsourced.
← All articlesEngineering
Application support SLAs: what to include and what to expect
How to define support tiers, severity levels, response and resolution times, monitoring and reporting when you outsource application support and maintenance.
Once an application is live, keeping it healthy becomes a job of its own. Many companies outsource this to a specialist team — and the service level agreement (SLA) is what turns “we’ll take care of it” into a measurable commitment.
Support tiers
- L1 — first response. Triage incoming issues, answer known questions, follow runbooks.
- L2 — application support. Investigate logs and data, fix configuration problems, apply workarounds.
- L3 — engineering. Code changes, root-cause analysis and permanent fixes.
Make sure your partner covers all three, or define clearly who owns each.
Severity levels
A simple, shared definition avoids arguments at 2 a.m.:
| Severity | Example | Typical target response |
|---|---|---|
| S1 — Critical | Production down, payments failing, data at risk | Within 30–60 minutes |
| S2 — High | Major feature broken, no workaround | Within a few hours |
| S3 — Medium | Feature impaired, workaround exists | Within one business day |
| S4 — Low | Cosmetic issue or question | Within a few business days |
Agree on response time (someone is working on it) and resolution or workaround time separately.
Coverage hours
Decide which hours need live coverage. A team in India naturally covers the UK and European working day and US early mornings and evenings; 24/7 coverage for critical systems can be added with on-call rotations.
What else belongs in the SLA
- Monitoring and alerting — uptime, errors and performance, with clear alert routing.
- Security updates — how quickly patches for critical vulnerabilities are applied.
- Backups and recovery — backup frequency, retention and tested restore times.
- Change management — how fixes are tested and released.
- Reporting — a monthly report on tickets, SLA performance, incidents and recommendations.
- Enhancement hours — a pool of time for small improvements, so support is not only reactive.
Handover done right
When a new team takes over support, insist on a structured knowledge-transfer phase: architecture walk-throughs, access set-up, runbooks and a period of shadowing before the SLA starts.
Choosing a partner
Look for a team that operates software of its own and understands production. At AkMo we support our own products every day — and offer application support and maintenance with clear SLAs for client applications. Talk to us about your support needs.