The 10 essential metrics for help desk software fall into two groups. Operational metrics measure the service: SLA compliance, customer satisfaction, tickets created outside the portal, new tickets, contract hours used and tickets in progress. Productivity metrics measure the team’s work: resolution time, tickets resolved, tickets that require training and agent assignment. To be useful, each metric needs a clear definition, a way of calculating it, a criterion for interpreting it and an associated decision. The sections below explain all ten.
Why measure help desk performance?
Without metrics, the perception of the service depends on complaints or on the team’s impressions. With them, it is possible to know whether agreed deadlines are being met, where work is piling up, which topics generate the most requests and whether users are satisfied. Measuring makes it possible to move from reacting to problems to anticipating them, and to back decisions on resources, processes or training with data.
What is the difference between operational and productivity metrics?
- Operational metrics: measure the service the user or customer receives: deadlines, satisfaction, channels and request volume.
- Productivity metrics: measure how the support team works: how long it takes, how much it resolves and how the workload is distributed.
Both are necessary. A team can be very productive and still miss its SLAs because it lacks resources. And the other way around: a service can meet its deadlines at the cost of overloading a few agents.
How do you choose the right metrics for your organization?
There is no need to measure everything from the start. It is best to begin with the metrics that reflect the service’s goals, for example meeting SLAs with customers, reducing repetitive tickets or balancing the team’s workload, and add others as new questions arise. Each metric should have an owner who reviews it and an associated decision: if a metric does not change any decision, it is unnecessary.

Operational help desk metrics
1. SLA compliance
- What it measures: the percentage of tickets handled within the response and resolution deadlines set by the SLAs.
- How it is calculated: tickets within deadline ÷ tickets with an SLA × 100.
- How to interpret it: if it drops, the deadlines are unrealistic or capacity is lacking. It should be analyzed by priority, group and customer, because an overall average can hide a group that almost always misses its deadlines.
- What decision it supports: reallocating resources, adjusting SLAs or reviewing agreements with the internal teams and suppliers involved. When an SLA is breached, the steps to restore it are in what to do if an SLA is breached.
2. Customer satisfaction
- What it measures: how users or customers rate the service they received.
- How it is calculated: the average score of the surveys sent when tickets are closed. It can also be expressed as the percentage of positive responses.
- How to interpret it: it should be read together with the number of responses, since a small number is not representative, and with the comments. It is more useful by category or by agent than as an overall figure.
- What decision it supports: training the team or reviewing the processes in the lowest-rated categories. Different types of satisfaction surveys can be used, such as star-rating surveys.
3. Tickets created outside the user portal
- What it measures: how many requests come in through channels other than the user portal, such as email, phone or chat.
- How it is calculated: tickets from other channels ÷ total tickets × 100.
- How to interpret it: a high value indicates that users are not using self-service, because they do not know about it, cannot find what they are looking for or find other channels faster.
- What decision it supports: promoting the portal, improving its forms and expanding the knowledge base in the topics that most often come in through other channels.
4. New tickets
- What it measures: the volume of requests coming in during a period.
- How it is calculated: the number of tickets created per day, week or month, by category and by channel.
- How to interpret it: peaks and trends show when and in which topics demand is growing. A sudden increase in one category usually points to an incident affecting many users.
- What decision it supports: sizing the number of agents and shifts, and detecting widespread incidents or underlying problems early.
5. Contract hours used
- What it measures: how many hours of a support or maintenance contract have been used by the requests linked to it.
- How it is calculated: hours logged on the contract’s requests ÷ contracted hours × 100.
- How to interpret it: it should be compared with how much of the contract period has elapsed. If 80% of the hours have been used halfway through the year, the contract will run out early.
- What decision it supports: notifying the customer in advance and proposing an extension or renewal of the contract.
6. Tickets in progress
- What it measures: how many tickets are still open and what status they are in.
- How it is calculated: the number of open tickets by status (assigned, in progress, pending user…) and by age.
- How to interpret it: if tickets pile up in one status, there is a bottleneck. Older tickets are usually the ones most at risk of breaching their SLA.
- What decision it supports: redistributing the workload, reviewing the oldest tickets and identifying those waiting on a third party.
Support team productivity metrics
7. Resolution time
- What it measures: how long it takes to resolve a ticket.
- How it is calculated: in two ways, depending on what you want to know:
- Total time: sum of (resolution date − creation date) ÷ resolved tickets. It measures how long the user waits.
- Time worked: sum of the time logged in the ticket’s actions. It measures the team’s effort.
- How to interpret it: it should be analyzed by category and priority. If total time is much higher than time worked, tickets spend a long time waiting.
- What decision it supports: documenting the most common solutions, training the team or automating the most time-consuming tasks. Reports show the breakdown by agent and category.
8. Number of tickets resolved
- What it measures: the team’s resolution capacity over a period.
- How it is calculated: the number of tickets resolved in the period. To interpret it, compare it with tickets created: tickets resolved ÷ tickets created.
- How to interpret it: if the ratio is below 1, more tickets are coming in than are being resolved, and the backlog is growing.
- What decision it supports: reinforcing the team during peak periods or reducing incoming tickets through self-service.
9. Tickets that require training
- What it measures: requests that are not real incidents, but questions that could be avoided if users knew how to do something.
- How it is calculated: tickets marked with a custom “training” field ÷ total tickets × 100.
- How to interpret it: a high value for a topic, application or customer points to a lack of training or documentation.
- What decision it supports: organizing training sessions, offering training to the customer or creating knowledge base articles. This frees up agents for real incidents.
10. Agent assignment
- What it measures: how the workload is distributed among agents.
- How it is calculated: the number of open tickets assigned to each agent and, as a complement, tickets resolved per agent.
- How to interpret it: an uneven distribution overloads some agents and leaves others with spare capacity, which lengthens resolution times and puts SLAs at risk.
- What decision it supports: assigning new tickets to the least busy agent and reviewing assignment rules. How to do this is explained in dashboards for a complete picture.
Summary: the 10 metrics at a glance
| Metric | How it is calculated | What decision it supports |
|---|---|---|
| 1. SLA compliance | Tickets on time ÷ tickets with an SLA × 100 | Reallocate resources or adjust SLAs |
| 2. Customer satisfaction | Average of closing surveys | Train or review low-rated processes |
| 3. Tickets outside the portal | Tickets from other channels ÷ total × 100 | Promote the portal and the knowledge base |
| 4. New tickets | Tickets created per period | Size agents and shifts |
| 5. Contract hours used | Hours used ÷ contracted hours × 100 | Extend or renew the contract |
| 6. Tickets in progress | Open tickets by status and age | Redistribute the workload and review old tickets |
| 7. Resolution time | Total time or time worked ÷ resolved tickets | Document, train or automate |
| 8. Tickets resolved | Resolved ÷ created | Reinforce the team or reduce incoming tickets |
| 9. Tickets that require training | Training tickets ÷ total × 100 | Organize training or create articles |
| 10. Agent assignment | Open tickets per agent | Balance the workload and review assignment rules |
How can these metrics be visualized?
Calculating them by hand is not practical: to be useful, metrics need to be available in real time and within reach of whoever has to make the decision. There are three ways to do this:
- Dashboards: show metrics in real time, with charts and alerts.
- Scheduled reports: sent to managers at regular intervals, for example every Monday.
- Custom views: ticket lists filtered by status, agent or SLA deadline, for day-to-day work. ServiceTonic makes it possible to create custom views.
What sets ServiceTonic apart in help desk measurement?
ServiceTonic makes it possible to measure all these metrics in custom dashboards that each manager configures without programming, with the indicators, filters and charts they need. The same data feeds scheduled reports and team views, so management, supervisors and agents all work with the same figures. Because the platform records every ticket from intake to closure, with its channel, SLA, actions and survey, all metrics come from the help desk software data itself, without separate spreadsheets.
Frequently asked questions
What are help desk metrics?
They are indicators that measure how the support service is performing: whether deadlines are met, how long it takes to resolve requests, how many requests come in, how work is distributed and whether users are satisfied. They help detect problems and make data-driven decisions.
What are the most important help desk metrics?
It depends on the service’s goals, but the most common are SLA compliance, resolution time, customer satisfaction, the volume of new tickets and workload per agent. Together, they give a view of both the service and the team.
What is the difference between a metric and a KPI?
A metric is any data point that is measured, such as the number of tickets created. A KPI (key performance indicator) is a metric tied to a specific goal, such as meeting 95% of SLAs. All KPIs are metrics, but not all metrics are KPIs.
How often should metrics be reviewed?
Operational metrics, such as tickets in progress or SLAs about to expire, should be reviewed daily. Trend metrics, such as satisfaction, ticket volume or average resolution time, are best analyzed weekly or monthly.
How is the success of a help desk measured?
By combining service and team metrics: a help desk works well when it meets its SLAs, maintains high satisfaction, resolves at least as many tickets as it receives and distributes the workload evenly. No single metric is enough on its own.
Conclusion
A metric is only useful if someone reviews it and a decision depends on it. Measuring ten indicators that nobody checks does not improve the service; measuring three that are reviewed every week and lead to concrete changes does. That is why, before adding a metric to the dashboard, it is worth asking who will look at it and what they will do if it changes.
Would you like to see these metrics on a dashboard with your own data? Request a demo and see how to measure your help desk’s performance with ServiceTonic.
