For every service that is included in the scope, an objectively measured SLA is specified, with financial penalties for non compliance, which typically kick in after a three months burn in period post go-live. Typical SLA examples could be % uptime for systems, or Mean Time to Fix for application issues, or % of issues fixed on first touch by the Helpdesk.
Typically the contracts would also include quarter on quarter SLA improvement targets, and ticket number reduction targets, as the service should continuously improve and all businesses ideally want no tickets.
Every year, there will be a cost reduction target as well, reducing the ongoing monthly service charge.
The monthly service charge will not be dependent on ticket count, but on SLA compliance. If we hit the SLA targets the cost will be predictable, if we occasionally miss, the cost will be reduced thanks to the financial penalties. This renders predictability and budgeting easier for our customers. Obviously we are betting we hardly ever miss.