Freshdesk Groups and Agents, Designed Properly

Freshdesk groups and agents are the skeleton of a helpdesk. Get them right and routing is boring. Get them wrong and every rule you write is a patch.

Start free trial All guides

What Freshdesk groups and agents are for

A Freshdesk group is a set of agents plus the settings that hang off it: assignment behaviour, business hours, escalation when nobody picks a ticket up.

The mistake is treating groups as an org chart. They are not. A group should represent a queue of work that a defined set of people can handle, which is often not the same as a team on a diagram.

Concretely: "Billing" is a good group because billing tickets need billing knowledge. "EMEA Team" is usually a bad group, because it describes where people sit rather than what they can answer, and now you need a second dimension of rules to work out who gets what.

How many groups you need

Fewer than you think. Under ten agents, two or three is plenty.

General support, which takes everything by default.

Billing or accounts, because the knowledge and the permissions differ.

Technical or escalations, once you've people who genuinely handle the hard cases.

Add a group only when you can answer two questions: what specifically lands here that would be answered worse elsewhere, and who is accountable if this queue backs up. If either answer is fuzzy, you're creating a place for tickets to get stuck.

The failure state is fifteen groups, most with two agents, several with one, and a routing ruleset nobody dares change.

Agents, roles and scope

Two settings decide what an agent experiences day to day.

Role controls what they can do: reply, edit, delete, change automations, export. Start restrictive. The number of teams who have discovered a well-meaning agent deleting tickets is higher than you would guess.

Scope controls what they can see: all tickets, only their group tickets, or only tickets assigned to them. Group scope is the right default for most people. Personal scope sounds tidy and in practice it means an agent can't see the ticket their colleague is already handling, which is exactly how one customer gets different answers.

Freshdesk also distinguishes full agents from occasional or collaborator-style access, which matters when you have engineers who touch a handful of tickets a month. Check what your plan includes before you buy seats for people who log in weekly.

Round robin and the automatic options

Freshdesk offers automatic assignment within a group, including round robin distribution and, on higher plans, load-based and skills-based variants. Round robin hands each new ticket to the next available agent in turn.

It works well when tickets are broadly similar in size and every agent in the group can handle any of them. It works badly when ticket effort varies wildly, because the agent who drew three two-hour investigations is now three hours behind the agent who drew three password resets, and the algorithm doesn't know or care.

Two practical guards. Set the agent availability toggle so people who are off, in a meeting or on another channel are skipped rather than accumulating a silent backlog. And keep an unassigned overflow path, so a group where everyone is unavailable doesn't quietly swallow tickets.

Load-based assignment, where it's available, solves most of the fairness complaint. It's one of the more genuinely useful reasons to move up a tier.

Why routing to a group beats routing to a person

This is the rule I would put on the wall.

Assign to a group on arrival. Let a person claim it. Never route a customer request at a named human.

Named routing breaks in every ordinary circumstance. The person is on holiday, off sick, in a meeting, has changed role, or has left. The ticket sits assigned and untouched, and it doesn't appear in anyone unassigned view because technically it has an owner. Nobody notices until the customer chases.

Group routing degrades gracefully instead. The ticket is visible to everyone who can help, your unassigned count is a real number, and escalation rules have something meaningful to fire against.

The one exception is a genuine ongoing case, where continuity is worth more than availability. Even then, add a rule that reassigns to the group if the ticket goes untouched for a day.

The check nobody runs

Once a quarter, pull the count of tickets assigned to each agent that are still open and have had no agent activity for more than five days.

Every number above zero is a ticket someone believes is handled. It's the quietest failure in support, and it's almost always a symptom of routing at people instead of at queues.

FAQ

Frequently asked questions

How does Freshdesk ticket assignment work across groups?

Freshdesk ticket assignment routes to a group first, then to a person. Freshdesk round robin distributes within the group on the higher plans, and Freshdesk agent roles decide who is eligible to receive anything at all.

How many Freshdesk groups and agents does a small team need?

Two or three. General, billing, and technical if you have real specialists. More groups than that under ten agents creates places for tickets to hide.

Is round robin available on all plans?

Automatic round robin sits on the paid tiers, with load-based and skills-based assignment on the higher ones. The free plan expects manual assignment.

Should agents see all tickets or only their own?

Group scope for most people. Personal-only scope hides the ticket a colleague is already answering, which is how one customer receives two conflicting replies.

What happens to tickets assigned to a departed agent?

They stay assigned unless you reassign them. Before deactivating anyone, bulk-reassign their open tickets back to the group. Put it in the offboarding checklist.