Reports (under Insights in the sidebar) shows how your support is performing: how quickly customers get a first answer and every answer after that, how long tickets take to close, how much is coming in and where from, how each agent and desk is doing, and what customers think of it.
Every report uses the same definitions, so a number means the same thing wherever you see it. For exactly what each figure counts, see How IntelliDesk calculates your report numbers.
Filtering
Most reports have two controls at the top:
- Date range: Today, Last 7 days, Last 30 days, This month, Last month, This quarter, or your own dates picked on the calendar. The range always includes its last day, and Last 7 days means today plus the six days before it. With nothing picked, you see the last 30 days.
- Help desk: every desk together, or one desk on its own. Use this to compare teams or mailboxes.
The Monthly summary report picks a calendar month instead of a date range.
Who sees which desks:
- Organization admins see the Whole workspace: every desk, including desks they aren't a member of. A head of support doesn't need to join every desk to see all the numbers.
- Everyone else sees All my desks: the help desks they're a member of.
- If you're not an admin and not a member of any desk, Reports shows "You don't have access to reports for this workspace."
Export CSV and the weekly email
- Export CSV (top right of every report) downloads everything that report shows (the headline figures with the previous period, and each table and chart as rows) in one file you can open in Excel or Google Sheets. The file notes which desks and dates it covers. Times are in business minutes, so they add up and sort properly.
- Weekly email: switch it on for yourself and every Monday morning you get last week's figures (Monday to Sunday), compared with the week before. It covers the workspace or your desks, then each desk and each agent, with the same numbers attached as a CSV. It covers the same desks you can see in Reports. Send last week's now sends one straight away. Each person switches it on or off for themselves.
The KPI row and "vs previous period"
Each report opens with a row of headline figures (KPI tiles). Under each one is its change vs previous period: the same-length stretch of time just before your range. So if you pick 1–14 September, the comparison is with the 14 days before that.
- Counts and times show a percent change (for example +12%).
- Rates such as SLA % and CSAT show the change in percentage points (for example +3pp).
- Green means better and red means worse. For times and backlogs, going down is good, so a faster response shows green even though the number fell.
- No arrow appears when there's nothing to compare with, such as when the previous period had no tickets.
Hover over a tile's label to see exactly what it counts and what the previous period's figure was.
The reports
Monthly summary
A two-page report for one calendar month, with a written summary, comparisons with the month before, and a Download PDF button. See Your monthly summary report (and downloading it as a PDF).
Volume
KPI tiles: Tickets created, Tickets resolved, Median first response and Median resolution time.
The Volume trend chart shows tickets created and resolved side by side as bars, Daily or Weekly. Weeks run Monday to Sunday; the first and last week can be shorter if your range starts or ends mid-week, and the tooltip says how many days a week covers.
Below the trend:
- By priority
- By channel: where tickets came in: Email, Website form, Portal, Chat, WhatsApp, API, or Logged by agent.
- By helpdesk: tickets created per desk over time.
- Volume heatmap: when tickets arrive, by day of the week and hour, in your workspace's time zone. Useful for planning shifts.
Agents: the agent scorecard
KPI tiles: Resolved, Median first response, Median reply time, Replies sent, CSAT and First-contact resolution.
Workload distribution shows how many unresolved tickets each agent holds right now, so you can see who is overloaded. Below it, the Agent performance table has one row per agent who did anything in the period. Click an agent to open their scorecard: all their numbers at a glance, plus the tickets they closed per day.
- Received: tickets assigned to them in the period, however they were assigned.
- Open now: unresolved tickets assigned to them right now.
- Replies: public replies they wrote in the period. AI and automatic replies aren't counted.
- Median reply: how long customers waited for their replies, for every exchange in a conversation, not only the first. The wait runs from the customer's message (the first one, if they wrote several) to the agent's reply.
- Resolved: tickets closed in the period.
- Median 1st response: median business time to the first public reply, on tickets created in the period.
- Median resolution: median business time from created to closed, on tickets resolved in the period.
- SLA met: of the SLA targets decided in the period on their tickets (first response and resolution), the share met. Hover to see how many targets it's based on.
- CSAT: the share of 4–5 star ratings on surveys sent in the period. Hover to see how many ratings it's based on.
- Reopened: the percentage of the agent's resolved tickets that had been reopened at least once.
- First-contact resolution: the percentage of resolved tickets where the customer never had to write again after the first reply, and which were never reopened.
- Streak: consecutive workdays (Monday to Friday) with at least one ticket closed. Weekends don't break it.
Who gets the credit: replies and reply time go to whoever actually wrote the reply. Resolved, first response, SLA and CSAT go to the ticket's current assignee.
SLA
KPI tiles: First response met, Resolution met and Breaches.
SLA percentages only count targets that have an outcome: met, or missed. A ticket whose target is still running doesn't count either way, so a busy week isn't made to look worse just because some targets aren't due yet.
- Breakdown: first response and resolution percentages, then By priority and By helpdesk (the share of SLAs that finished with every target met).
- Compliance trend: the weekly share, by the week each outcome was decided. The first and last week may be partial.
- Breached in this period: tickets created in the period that missed any target, open or closed, showing which target each one missed. Click a ticket number to open it.
SLA targets are set in Settings → SLA Policies (see SLA policies and business hours).
Customers
KPI tiles: Contacts with tickets, Companies with tickets, Repeat contact rate (contacts who raised two or more tickets in the period) and Frustrated customers (tickets AI triage flagged as coming from a frustrated customer).
Below that is the Customer health table, one row per company, and Top contacts, the contacts with the most tickets in the period. See Customer health report: which customers need attention?
CSAT
KPI tiles: CSAT score (the share of ratings that were 4 or 5 stars), Avg rating out of 5, Response rate (the share of surveys that got a rating), Surveys sent, and Bad ratings (1–3 stars).
Below that: the CSAT trend by week, CSAT by helpdesk, CSAT by agent, and Recent bad ratings with the ticket each came from, so you can follow up. All of these count surveys sent in the period.
Surveys are sent when a ticket is closed, on desks where the satisfaction survey is switched on in the desk's settings. A customer is surveyed at most once every 7 days, so people who write in often aren't pestered.
Live chat
How many chats started, how many the AI answered alone, how many were handed to a person or moved to a ticket, how long visitors waited for a person, and chat CSAT. See Live chat: setting it up and answering chats.
Backlog
KPI tiles: Unresolved, Unassigned, Older than 7 days and Reopen rate.
- Unresolved is every ticket not closed right now: Open, Customer responded, Pending and On Hold. Its change is shown vs start of period, meaning the backlog as it stood when your date range began.
- Older than 7 days counts unresolved tickets in the 7–14 days and 14+ days buckets, also compared with the start of the period.
- Reopen rate is the share of tickets resolved in the period that had been reopened at least once.
The charts group unresolved tickets by age: under 1 day, 1–3 days, 3–7 days, 7–14 days and 14+ days, with a separate chart for Unassigned tickets. Ages are measured in business hours, where a "day" is 8 business hours. The Aging trend over time shows the backlog at the start of your range, each Monday, and today, so you can see whether old tickets are piling up. The backlog is the best early warning that a queue is falling behind.
AI & automation
What Jev, IntelliDesk's AI assistant, did in the period: Tickets triaged by Jev, Chats resolved by AI alone, AI replies sent and Assigned by best match, with a breakdown in What Jev did (priorities set, spam filtered, frustrated customers flagged, likely duplicates spotted and so on).
At the top is Agent time saved by AI (estimate). It is an estimate, not a measurement, and the page shows the assumptions behind it: 1 minute per ticket triaged, 1 minute per spam or machine email filtered, 6 minutes per chat the AI resolved alone, 4 minutes per AI first reply, and half a minute per ticket assigned. Every other number on the page is a count.
If Jev isn't switched on, the page shows "No AI activity in this period". Turn on AI triage & assist for a desk in Settings → Help Desks. For what Jev does and how to set it up, see What is Jev? The AI in IntelliDesk explained.
How times are measured
- First response runs from when the ticket was created to the first public reply from an agent. Internal notes don't count.
- Reply time covers every answer, not only the first: each time a customer writes and waits, the clock runs until an agent replies publicly.
- Resolution time runs from when the ticket was created to when it was closed.
- The headline figures are medians, not averages: half of tickets were faster, half slower. One ticket that sat for three weeks can't drag the whole number up.
- Times count business time only, using your workspace's default business hours (Settings → Business Hours). A ticket that arrives at 5pm on a Friday and is answered at 9am on Monday counts as minutes, not a whole weekend. If you haven't set business hours, the clock runs around the clock. Live chat replies often take seconds, so an agent who mostly chats can show a reply time of 0m.
What's left out
- Trashed tickets.
- Merged tickets. Merging closes the duplicate, which would otherwise count as an extra resolution.
- Live chats the AI is still handling on its own. They never reached your team, so they'd distort ticket numbers. The Live chat and AI & automation reports cover them.
Leaving a desk out of the numbers
Internal, test or system-log desks can skew your averages. In that desk's settings, turn on Hide from dashboards & counters and it's left out of the reports' all-desks view, Home and Overview. You can still report on it by picking it in the Help desk filter.
The Overview page
For a quick picture without choosing a report, open Insights → Overview in the sidebar. It shows:
- Unresolved, Open, Median first response, Median resolution time (last 30 days, compared with the 30 days before) and SLA breached now.
- Status and desk breakdowns, the desk backlog and top tags.
- The volume trend, response-time trend, agent leaderboard, workload, backlog aging and SLA compliance for the last 30 days.
If Overview, Home and Reports seem to disagree, see Why don't the numbers on Home and Reports match?