Email Productivity: Comments per Online Hour and per Open-Status Hour
How productive is an email agent? Counting replies alone rewards whoever worked the longest shift. Qvasa measures email output against the time it took, two ways:
Per online hour: replies, internal notes and tickets touched for every hour the agent was clocked in on the Email Support channel in Zendesk. This answers "what did we get for the hours we staffed?"
Per Open-status hour: public replies for every hour a ticket sat in an Open status while the agent held it. This answers "how quickly does this agent move the tickets they own?", and it works whether or not your team uses Zendesk's online status.
This article explains both families, exactly how each number is calculated, and how to choose between them.
Before you start
Per online hour metrics need agent status tracking on your account: Qvasa reading your agents' Zendesk availability, so it knows when each agent was online on the Email Support channel. They only have something to measure if your agents set themselves online for email in Zendesk.
Per Open-status hour metrics need Advanced Zendesk Ticket Comment metrics (ticket comment records) on your account.
Total public comments by agent needs Zendesk ticket audits.
If a metric below doesn't appear in your widget picker, ask your Qvasa contact to turn on the setting it needs.
Per online hour: output for the hours you staffed
Every time an agent goes online on the Email Support channel in Zendesk, Qvasa opens an online window for them, and closes it when they go offline or away. As the agent's comments sync, Qvasa counts the ones they made inside that window and stores the counts on it: public replies, private internal notes, and the number of distinct tickets they commented on.
The metrics
Metric | What it measures |
Total Email Support Online Hours | Hours agents were online on the Email Support channel: the denominator of every rate below. |
Total Comments / Total Public Comments / Total Private Comments While Email Support Online | Agent comments made inside online windows: all, public replies only, or internal notes only. |
Comments / Public Comments / Private Comments per Email Support Online Hour | Those comment totals divided by the online hours. |
Unique Tickets With Any / Public / Private Agent Comment per Email Support Online Hour | Distinct tickets the agent commented on, per online hour. A ticket with five replies counts once. |
Each one is available as a quick stat, an over-time chart, an agents table column and donut chart (by agent), and a by-agent over-time chart. Public Comments per Email Support Online Hour can also be broken down by Zendesk skill, by Zendesk user field, and in a metric matrix by user field option and skill, on accounts that sync those.
Add the columns to an Agents Table to compare agents side by side. Here Priya worked half the hours Maya did and wrote half the replies, so the two have the same rate. Leo writes as many internal notes as replies: his all-comments rate looks strong, but he touches fewer tickets an hour. Sam was online a full week and replied about twice an hour.
How it's calculated
A comment counts when it lands inside an online window. Qvasa matches by time: any comment the agent makes while online on Email Support counts, on any ticket. A comment made while the agent is offline counts in none of these metrics.
A window belongs to the day it started. The date range picks online windows by when the agent went online, and each window brings its whole length and all its comments. An overnight shift counts on the day it began. A window that is still open counts its hours up to now.
Rates are a ratio of totals. Every rate is total comments divided by total online hours across the range, never an average of per-agent or per-day rates. A one-hour shift can't swing the team number the way a full week does.
Per Open-status hour: output for the time spent on the ticket
These metrics don't look at online status at all. They're built on assignment sessions: each stretch during which one agent held one ticket. Inside a session, Qvasa records how long the ticket was in an Open status and how many public replies that agent sent while they held it. Time the ticket spent Pending or On-hold, waiting on the customer, never counts against the agent.
The metrics
Metric | What it measures |
Public Agent Total Ticket Comments Count (quick stat), Count of Public Comments by Agent (agents table column "Comment Count", and a donut) | Every public reply agents made in the date range, whether or not they were online. |
Email Average Public Agent Comments per Open-Status Hour | For each email session that ended in the range: public replies divided by the hours the ticket was in Open. One reply in a 30-minute Open stretch is 2.0 per hour. The metric is the average of those per-session rates. |
Email Median Public Agent Comments per Open-Status Hour | The median of the same per-session rates. Prefer it when a few unusual sessions pull the average around. |
Email Average / Median Assignment Session Open-Status Duration | How long tickets sat in Open per session: the denominator, shown on its own. |
The two rate metrics are available as a quick stat, an over-time chart, an agents table column and donut chart, and a by-agent over-time chart. Every widget has a drill-in to the sessions and agents behind it.
Per agent, the story lines up with the online view, with one difference worth noticing. Nora's total public comments (324) are higher than her public comments while online (270, in the table above), because she answers some tickets after going offline. Those replies count here and nowhere in the per-online-hour metrics. Sam's tickets stay Open about 50 minutes per session, and he averages under two replies per Open hour.
How it's calculated
Only Email-channel sessions count: email, web form and API tickets, by the channel the ticket was on during the session.
Sessions are picked by when they ended. A session still running isn't counted until it finishes.
A reply credits the agent who held the ticket. A reply sent after the ticket was handed to someone else belongs to nobody's session.
Short sessions are left out. By default a session needs at least 5 minutes in Open to count, so a ticket opened for seconds can't post an extreme rate. Change the minimum in the widget's settings; it shows in the widget title.
Sessions with no reply still count, as zero.
Public replies only. Internal notes aren't part of these rates.
Going forward from when stamping began. Sessions recorded before Qvasa started stamping comments and Open time on them aren't included, and there's no backfill.
Which one to use
Question | Metric |
Are we getting enough output for the hours we staff? | Public Comments per Email Support Online Hour |
Who is clocked in but not working tickets? | Unique Tickets With Any Agent Comment per Email Support Online Hour, by agent |
How much internal back-and-forth is there? | Private Comments per Email Support Online Hour next to the public rate |
How fast does an agent move the tickets they own? | Email Median Public Agent Comments per Open-Status Hour |
My team doesn't set itself online in Zendesk | Public comments by agent and the per Open-status hour metrics |
Comments or tickets? Comment counts reward many short replies, and internal notes inflate the all-comments rate. Unique tickets per hour is the fairer measure of throughput: it counts each ticket an agent worked once, however many comments it took.
Good to know
An agent who never goes online on Email Support shows no online hours and no per-online-hour numbers, however much they reply. Use the Open-status metrics for those teams.
Online status is the agent's own setting. An agent who forgets to go offline adds idle hours and lowers their rate; one who works without going online adds output nobody sees in the online view. A big gap between an agent's total public comments and their public comments while online usually means one of these.
Breakdowns by skill and user field are for public comments per online hour. The other rates break down by agent.
Related articles
Agent handle time by channel with assignment session open-status duration: the rest of the assignment session metrics, including time to first reply
Open-status metrics on ticket assignment sessions: every session metric, live and historical
Understanding email touch count metrics: replies per solved ticket
Measuring agent idle time on Messaging: the Messaging view of online time