At the end of the month, the SLA report arrives with several breaches. The team pushes back: that incident came in on Friday evening and the service is only supported during business hours. This one waited three days for the customer to reply. And that one fell under a different contract with different targets. The operations manager spends hours reconstructing what actually happened, while the customer has already received figures that do not reflect what was agreed.
Badly measured SLAs do damage in two ways. A team that did good work is reported as failing and loses motivation. Meanwhile, real problems get buried among false breaches and nobody addresses them. Trust in the numbers erodes on both sides.
Why simple measurement does not work
The simplest approach is to count the time from ticket creation to resolution. But in practice an SLA rarely means “24 hours, around the clock”. It usually means “within one business day, during supported hours”. That is a very different thing.
We often see the same mistakes:
- the clock runs overnight, at weekends and on public holidays, even though the agreement only covers business hours,
- time spent waiting for the customer or a supplier is counted as resolution time,
- every incident is measured against one default target, even though different services and customers have different agreements,
- nobody is warned about an approaching breach, so it only surfaces in the monthly report.
How to approach it
SLA measurement should mirror the agreement exactly. That means a few things. First, every target has to know which calendar it runs on: what the business hours are and which days are holidays. Second, there must be an honest way to pause the clock when work is genuinely blocked outside the team, and that state must be visible. Third, agreements form a hierarchy: a specific agreement for a service or customer takes precedence over the default.
Finally, measurement should not only judge the past. It should help people meet the target while there is still time to do so.
How Eywo Stratum handles it
In Eywo Stratum, incidents and requests carry response and resolution SLAs calculated in business hours. Business hours are defined by calendars that include public holidays, so a target expressed in working hours is not consumed over a weekend or a bank holiday.
Other elements of measurement:
- pausing – when resolution is waiting on something outside the team, such as the customer, the clock can be paused, and the state is visible on the ticket,
- SLA agreements – targets can be set for a specific service, for a customer, and as a default, so each incident is measured against the agreement that actually applies to it,
- priority from impact and urgency – the target reflects real severity rather than the impression of whoever opened the ticket,
- monthly SLA performance – an overview of how agreements are being met over time, for internal management and for conversations with customers.
Warnings ahead of a breach come through email and Slack notifications driven by rules and templates. The resolver and team lead learn that a deadline is approaching while there is still time to act. For incidents that depend on a supplier, escalation follows the service contract.
Because an incident is raised from a CMDB item and the system suggests the affected service, it is clear from the start which SLA agreement applies. Nobody has to dig out the contract after the month has closed.
Numbers you can trust
A well-measured SLA is not a tool for finding someone to blame. It is a shared language between operations and the people who rely on its services. When the numbers match the agreement, the month-end report becomes the basis for a conversation about improvement, not an argument about methodology. And a team that did good work can see that reflected in the figures.
SLA agreements, business-hour calendars, pausing and monthly SLA performance are included in the Standard edition.
Eywo Stratum Standard
For organisations that want a reliable CMDB and managed ITSM processes.