Ticket routing is arguably one of the most important components of how Zendesk works. From Zendesk's own documentation:
One of the best ways to increase agent efficiency and streamline your support tasks is to use Zendesk routing options to manage your ticket workflows. Routing options provide a power set of tools to make sure your tickets get to the right agent as quickly as possible.
With it being so important, understanding the ins and outs of how tickets route from creation to being worked is vital to truly understanding how GitLab uses Zendesk as a whole.
As this can get complex, Support Ops breaks down the routing into stages. These stages are not meant to say one flows into the next into the next. Instead, they help to break down the complexity into more digestable chunks. Some of these can happen at the same time, while others depend on previous stages to complete before they can run.
This is where tickets all start. This is what happens immeditaly once tickets are created.
This stage focuses around associating an end-user to an organization.
In this diagram, the zapier process goes as follows:
Zapier searches SFDC for the contact
This stage is currently a placeholder. The plan is to have this stage cover missing metadata, however for the time being, this stage is skipped.
This stage covers setting the Area of Focus on tickets.
This stage handles the big routing steps, namely priority and schedule. These two fields will determine the initial SLA on a ticket. A big not here is this stage is not linear. The steps for priority are listed first, but that doesn't mean a hinderance there would block the schedule steps. The flowchart refers purely to their position in Zendesk.
There are a few stages remaining, but they have more to do with the lifespan of a ticket than its actual routing. For more information, see the Triggers.