Service Level Agreement
The standard framework GIPSIC starts from for support contracts. The real numbers are set in each engagement's own contract.
Effective 15 July 2026
This is the framework we start negotiations from, not a commitment in itself. The service levels that actually apply are the ones written into your contract or quotation. Where the two differ, the contract governs.
Scope
Applies only to systems under a signed support and maintenance contract. It does not cover new development, changes of scope, or systems whose support has ended.
Severity
| Level | Meaning | Example |
|---|---|---|
| P1 Critical | The system is unusable, or revenue-bearing transactions are directly affected, with no workaround. | Nobody can sign in; no payment completes. |
| P2 High | A core function is unavailable, but a workaround exists. | Monthly reporting fails, though raw data can still be extracted. |
| P3 Medium | Usable, but wrong or awkward. | A summary figure is off; part of a screen renders incorrectly. |
| P4 Low | Observations, small requests, and questions about use. | Unclear button wording; a request for another report filter. |
Severity is agreed between the person raising the incident and our team. Where there is disagreement, the higher level applies until it can be assessed.
Response and resolution
| Level | Acknowledged within | Work begins within | Restoration target |
|---|---|---|---|
| P1 | as stated in the contract | as stated in the contract | as stated in the contract |
| P2 | as stated in the contract | as stated in the contract | as stated in the contract |
| P3 | as stated in the contract | as stated in the contract | Next release |
| P4 | as stated in the contract | As scheduled | As agreed |
Response time is when we acknowledge the incident and name an owner. It is not when the problem ends. We do not commit to a closure time for a fault whose cause is not yet known.
Service hours
- Normal channels: the working days and hours stated in the contract.
- Emergency channel for P1: as stated in the contract, which also says whether it covers out of hours.
Availability
Monthly availability target: as stated in the contract. Calculated as the time the system was serving requests, divided by the total time in the month, less the exclusions below.
Where a system runs on third-party infrastructure, the availability we can commit to is never higher than what that provider commits to us.
Raising an incident
Use the channel named in your contract, and please include:
- what happened, and what you expected instead;
- when it was first seen;
- who or which department is affected, and roughly how many people;
- steps to reproduce it, if you have them;
- screenshots or error codes.
The response clock starts once we have enough to begin investigating.
Planned maintenance
We give at least the notice period stated in the contract and choose the window that disturbs the fewest users. Announced planned downtime does not count against the availability calculation.
Exclusions
This service level does not cover:
- force majeure, and outages at network or cloud providers outside our control;
- changes the client or another supplier makes to the system, its data or its environment without telling us;
- use beyond the scope or volume the system was designed for;
- faults in third-party software or hardware we did not supply;
- systems whose fees are overdue.
Service credits
Remedies for a missed service level are those set out in the contract for that engagement. Where the contract is silent, the parties discuss it case by case.
Review
The parties review this service level together at least annually, or whenever the scope of the system changes materially.
Contact
[email protected], or the channel named in your contract.