The Manager's Responsibilities for a Help Desk Team
Friday, 18:20, at Acme. Priyanka has run the desk of four for six weeks, having been its best technician for three years. She is finishing the monthly report. The reporting view in front of her shows tickets closed per technician, sorted from most to fewest. She is about to paste it in, because it is a table and tables look like management. At the bottom of the table is Theo, who closed the fewest tickets this month because he took every hard one. Those included the three-day certificate problem that nobody else could have fixed. Priyanka knows this. The table does not. She has also, this week, sent the team three e-mails after ten at night. She has not yet noticed that the team now reads them at five past ten.
The manager's responsibilities in this subject are the things only the person with authority can do to keep a desk sustainable. That authority covers staffing, the rota, the reporting and their own behaviour. The earlier articles in our Helpdesk Burnout series described the mechanisms: the interruption load, the queue-duty rotation, the recovery fortnight. Technicians can run most of those among themselves. What they cannot do is add a person, cut the scope, decide what goes in the report, or leave on time in a way that gives everyone else permission to. Those are the manager's. A desk where the manager does not do them is a desk where the mechanisms slowly fail, however well the technicians run them.
This article covers why the job is the load rather than the queue, then staffing honestly against the volume, with a worked example. Protecting the team from the queue's infinity and the one-to-one that asks and listens follow. Next come metrics that do not punish, modelling the behaviour, and noticing early. The manager's own load, a short war story, and the one-to-one check-in questions to copy follow, in that order.
The Job Is the Load, Not the Queue
A new desk manager, especially one promoted from the desk, usually believes their job is the queue. That means getting it down, keeping it down, and answering for it when it is up. But that is not the job. The queue is infinite in the sense that matters, which is that for every ticket closed another arrives. A manager who takes the queue as their personal responsibility has taken on a thing that cannot be finished and will take the manager with it. The job is the load: how much work the desk is asked to carry, how it is shared, and whether the people carrying it can keep doing so next year.
That reframing changes what the manager spends their time on. A queue manager spends the day in the queue, taking tickets, chasing tickets, asking about tickets. A load manager spends the day on five levers that nobody else on the desk holds. Staffing, meaning the number of people and their hours. Scope, meaning what the desk agrees to do and what it says no to. The rota, meaning how the load is divided across people and across the day. Reporting, meaning what the organisation is told about the desk and in what terms. And their own behaviour, which the desk copies whether or not the manager intends it. The rest of this article takes those in turn, more or less.
The hardest part for a promoted technician is that the levers are slower than tickets. A ticket closed is a thing done by lunchtime. A staffing case is a thing done by the end of the quarter, perhaps, if the arithmetic is honest and the timing is right. The temptation to go back to the queue, where progress is visible, is strong. Most new managers give in to it for about six months. I did. The queue was no smaller for it and the levers went unpulled, which is the usual result.
Staffing Honestly Against the Volume
Staffing honestly means putting the hours the work needs beside the hours the desk has, in writing, and saying out loud what the gap is. Most desks have never done this, which is why most desks are staffed by history. The number of people is whatever it was when the last person was hired, plus or minus whoever has left. The volume has moved since without anyone comparing the two. The comparison is a morning's work with the volume report and the one-week interruption log. That log comes from The Interruption Load and Why It Is Different. The table shows Priyanka's version for Acme, as a worked example from a fictional desk and nothing more.
| Acme desk of four (worked example) | Hours a week | Where the figure came from |
|---|---|---|
| Tickets arriving | About 210 tickets | The volume report, averaged over three months |
| Handling time for those tickets | About 70 h | 210 tickets at a median of 20 minutes each, from the ticket data |
| Recovery time and unlogged work | 25 to 35 h | The one-week interruption log, scaled across four people |
| Admin, meetings, training, queue hygiene | 15 to 20 h | Calendars, plus the weekly hygiene routine |
| Hours the work needs | 110 to 125 h | The three rows above added together |
| Hours the desk has at the desk | About 112 h | Four people, less leave, absence and time away from the queue, averaged over the year |
| The gap | Between 2 h spare and 13 h short, every week | The desk is at its edge; any leave or absence tips it under |
The gap is the finding. There are only three honest things to do with a gap: add hours, cut scope, or remove work. Adding hours is the headcount case. The method for making it with the desk's own figures rather than with anecdote is in The Headcount Case. Cutting scope means deciding, visibly, that the desk no longer does some category of thing, with the organisation told. That is the subject of Scope Boundaries. Removing work means fixing the recurring problem, writing the article that answers the question before it is asked, or automating the reset. The whole deflection and prevention half of this library is about that. There is a fourth option, which is "work harder". It is not honest, because the table already contains everyone working as hard as they can.
Say the gap upward in those terms, with the table, before the desk is in trouble rather than after. "The work needs between a hundred and ten and a hundred and twenty-five hours a week and we have a hundred and twelve; here is where the figures come from, and here are the three things we could do about it." A director given that sentence has a decision to make. A director given "the team is stretched" has an opinion to form. Opinions about whether a team is really stretched tend to be formed in the director's favour. The arithmetic is also the manager's protection. When the answer is no, the manager has a dated record of having said what the gap was.
When the answer is no, cut scope visibly rather than letting the people absorb it invisibly. A desk that is thirteen hours short every week will find the thirteen hours somewhere. If the manager does not decide where, the technicians will, by staying late, skipping lunch and dropping the notes. That is the first article in this series happening in slow motion. The manager who says "we will stop doing X until the staffing changes, and I have told the organisation so" has turned an invisible cost on people into a visible cost on service. That is the only kind the organisation can see and therefore the only kind it will ever fix.
Protecting the Team From the Queue's Infinity
The queue count is the most visible number on the desk and the least meaningful one. A manager's first protective duty is to stop the team reading it as a verdict. Eighty open tickets is not a failure if the oldest is three days old and the desk is closing as many as arrive. Forty is not a success if a dozen of them are ninety days old and the requesters have given up. The numbers worth watching are the age of the oldest ticket, the count over a month old and whether the week's arrivals and closes balance. A manager who talks about those, and never about the raw count, teaches the desk to do the same.
The second protective duty is to absorb the pressure from above rather than passing it through. When the director asks why the queue is eighty, the manager answers, with the age figures and the arithmetic. The question does not reach the desk as "the director wants to know why the queue is eighty". A manager who forwards that question to the team has done nothing with it except add a director to the load. The manager's job, in that moment, is to be the place where the organisation's worry about the queue stops. That is uncomfortable, which is why it is the manager's job and not the technicians'.
The third is what goes on the wall. A screen that shows the count in large red digits, all day, to people who cannot change the arrival rate, is a device for making people feel bad about arithmetic. A screen that shows what is waiting, what is overdue and who holds what is a tool. The difference between the two, and how to design the second, is in Dashboards That Motivate Instead of Surveil. The manager chooses which one the desk looks at. Nobody has ever closed a ticket faster because the number above their head was redder.
The One-to-One That Asks and Listens
A one-to-one is a private, regular conversation between the manager and one technician, half an hour every fortnight or so, that is not about tickets. It is the only reliable way a manager finds out what the load feels like from inside. That is because the ticket data shows what people did and the one-to-one shows what it cost them. The two are different things. Most desks either do not have them or have them as a status update on open tickets, which is a different meeting with the same name.
The one-to-one has three parts, and the template at the end of this article is the questions for each. The load: what has been heaviest, what has been interrupting most, what did they not get to. The work: what is working well, what is broken in the process, what would they change if they could. And the person: how are they, honestly, and is there anything the manager should know. The order matters, because the third question asked first gets "fine". Asked third, after twenty minutes of talking about the work, it sometimes gets the truth.
Listening is the part that fails. A manager who has been a technician hears a problem and solves it, out loud, immediately. That ends the conversation and teaches the technician that the one-to-one is a place where problems get fixed rather than heard. Some should be fixed, later, by the manager, after the meeting. In the meeting, the job is to ask the next question. "What else?" is the most useful sentence in a one-to-one and the one most managers never get to. That is because they are already drafting the solution to the first thing.
Then do one thing. A one-to-one that produces a list of nine problems and no change teaches the technician that the meeting is theatre. They will stop bringing problems to it within three cycles. Pick one, say which, change it before the next meeting, and open the next meeting by saying what changed. The quiet technician described in What Burnout Actually Looks Like on a Help Desk is often quiet because they brought nine problems once and nothing happened. The one-to-one is where the manager earns the right to be told the tenth.
Remember: the one-to-one is not about tickets. Ask about the load, then the work, then the person, in that order. Listen longer than feels comfortable. Change one thing before the next one and say what it was.
Metrics That Do Not Punish
Every number that can be cut by technician will be read as a ranking by the first person who sees it sorted. A manager's job is to make sure that reading is wrong. Per-technician figures, closes, reopens, satisfaction, handling time, exist for one purpose on a healthy desk. That is a coaching conversation between the manager and one person about that person's own work, in private. There must be enough tickets behind the number for it to mean anything. Those figures do not go on the wall. They do not go in the monthly report by name. They do not get sorted. The rule about how many tickets a per-person figure needs before it is read at all is argued properly in Satisfaction by Category and Carefully by Agent. The manager adopts it rather than re-deriving it.
The reason is not kindness, or not only. A ranked close count rewards the person who takes the easy tickets and punishes the one who takes the hard ones. Within a month of the ranking going up the desk will have reorganised itself so that nobody takes the hard ones. That is how Theo ends up at the bottom of a table that will then be used to decide he is the weakest person on the desk. The metric does not measure performance. It measures what the desk learned to do about the metric, and the desk learns fast.
Used properly, the same numbers are the best coaching tool a manager has. A reopen rate that is higher for one person than for the desk, over a few hundred tickets, is a conversation. It is about which kinds of fix come back, with the tickets open on the screen. It usually ends with a runbook or a habit rather than a judgement. A close count that is low because the tickets were hard is a conversation about whether the hard tickets should be spread more evenly. If you run QuerySolv, the per-technician figures in the reports are there for exactly that conversation, between one manager and one technician. Nothing in those figures is improved by being shown to the rest of the room.
The honest version of the monthly report shows the desk's figures as a desk: volume in and out, age of the backlog, reopen rate, satisfaction. It includes a line about load where the load is the story. It names nobody. The shape of that report and what belongs in it is in The Numbers to Show and the Numbers to Keep. Priyanka's per-technician table belongs in the second category. Theo belongs in a one-to-one, being told that the certificate problem was good work.
Modelling the Behaviour
The desk does what the manager does, not what the manager says. This is the responsibility that promoted technicians most underestimate, because for three years nobody copied them. Now everybody does. A manager who leaves at half past five, visibly, with a "see you tomorrow", has given four people permission to leave at half past five. A manager who sends e-mails at ten at night has told four people that ten at night is when the desk works, whatever the policy says about leaving on time. The e-mails do not need to say "reply now". The timestamp says it for them.
The list is short and all of it is visible. Leave on time, and be seen leaving. Do not send anything to the team outside the desk's hours. Write it, hold it, send it in the morning. Take lunch away from the desk, so that lunch away from the desk is a thing people on this desk do. Book leave and take all of it, so that a full leave balance is normal rather than a sign of not caring. Hold a queue-duty block once a fortnight, as the rotation in Sustainable Pacing describes. That makes the rota something the manager is in rather than something the manager imposes. None of these is hard. Every one of them is undone by a single visible exception.
The exception the desk notices most is the manager who sets the rules and then breaks them "just this once" for a bad week. The desk does not hear "just this once". It hears that the rules apply when things are calm, which is to say when they are not needed. It hears that when things are hard the real rule is the old one. If the manager must work late in a genuine incident, the recovery fortnight in Recovery After a Bad Stretch applies to the manager too, visibly, including the day off. The desk watches whether the manager takes it. Then it decides whether the rule is real.
Noticing Early
The manager is the person best placed to notice the early signs and the person most likely to miss them. That is because the manager sees the desk through its numbers and the early signs are not in the numbers until late. Noticing early is therefore a routine rather than an instinct. It is the manager half of the early-signs checklist from the first article in this series. Run it for each person once a month on a calendar reminder, from what the manager has seen and from the ticket signals. Three ticks is a conversation. Three ticks two months running is action, this week, whatever the queue looks like.
Add the walk. Once a day, the manager stands up and goes round the desk with no question to ask and nothing to deliver, and looks. Who has headphones on all day. Who is eating at the keyboard. Who has stopped turning round when someone comes in. It takes four minutes, and it is the only instrument that notices the person who has gone quiet. That is because a quiet person generates no tickets, no e-mails and no numbers. They are therefore invisible to every other tool the manager has. The walk is not supervision. It is attention, and the difference is whether the manager is looking for problems or looking at people.
Where the signs are there, act in the order the first article gave: lighten the load first, then ask. A manager who notices the signs and schedules a one-to-one for next Thursday to discuss them has noticed and done nothing for a week. Move two duty blocks today, and have the conversation on Thursday about a load that is already lighter. The conversation goes better when the person can see that noticing led to something.
The Manager's Own Load
Managers burn out too, and the desk manager is unusually exposed. That is because they carry the desk's load as its shield and the organisation's worry as its filter. They usually carry some of the queue as well, since a desk of four cannot spare a whole person for management and the manager knows it. The signs are the same ones: the shortening replies, the late e-mails, the one-to-ones cancelled for the third time because something came up. The manager rarely notices them in themselves, for the same reason technicians do not, and the manager has fewer people watching.
The protections are also the same, applied upward. The manager has a one-to-one with their own manager. In it, they say what the load is, in the arithmetic above, and what they are doing about it. They do not cancel it because the queue is bad. The manager takes their day off after a stretch and their leave in the year, visibly. The manager holds queue duty once a fortnight and not every day. That is because a manager who is in the queue all day is a technician with a title and a bigger rota problem. Nobody is pulling the levers while they are there. And the manager delegates what can be delegated, which on a small desk is more than it looks. That includes the hygiene routine, the handover note, and the knowledge base review. A senior technician can own each of these, and would like to own most of them.
The last protection is the hardest. The manager accepts that the queue will not be zero, that some targets will be missed, and that the headcount case may be refused. The manager accepts that none of those is a personal failure. In the same way, the manager has spent this whole article teaching the desk that the count is not a verdict on them. The queue is infinite. The manager is not. Managing the gap between those two facts, for the team and for oneself, is the job. It is a job that gets done well in years rather than weeks.
A Short War Story: The Leaderboard at Larkfield Academy
At Larkfield Academy, a school with a desk of three, a new manager called Gordon put a screen on the wall. It showed tickets closed per technician, sorted, updated through the day, and he called it motivation. Within a month the desk had learned the screen. The two junior technicians took every password reset and printer jam the moment it arrived. The senior technician, Beatriz, had always taken the long problems with the student information system. She found herself permanently at the bottom of a list that the head of operations could see from the corridor. The reopen rate roughly doubled over the term, because tickets were being closed to move a number rather than to fix a thing. Beatriz, after a one-to-one in which Gordon asked why her figures were low, resigned at the end of the summer term. Replacing her took the whole of the autumn. During that time the student information system's two worst problems went unfixed. The head of operations asked why the desk's closed count had fallen, since the screen now showed it. Gordon took the screen down the week after she left and wrote in his notes that the number had never measured the work. It had measured what three people had learned to do about the number. The one who refused to learn it had been the one worth keeping.
The One-to-One Check-In Questions
The sheet below is the manager's side of a fortnightly one-to-one. It is a page with the questions in the order to ask them, space for what was heard, and three lines at the bottom. Those lines are the only part the manager fills in after the meeting. Ask the questions as written the first few times, until they stop sounding like a script, and never ask the third group first. The sheet is for the manager's own file, not for the technician's record, and the technician is told that. Keep last fortnight's sheet open during this fortnight's meeting, because the first thing to say is what changed.
ONE-TO-ONE CHECK-IN — FORTNIGHTLY, THIRTY MINUTES, NOT ABOUT TICKETS Technician: ______________ Date: ________ Last one-to-one: ________ OPENING (the manager speaks first) What changed since last time, from what you told me: _________________________ 1. THE LOAD What has been the heaviest part of the last two weeks? What interrupted you most? Where did it come from? What did you not get to that you wanted to? How did the queue-duty blocks feel? Too many, too few, wrong times? Heard: __________________________________________________________________ 2. THE WORK What is working well that we should keep? What in our process is broken, slow or pointless? If you could change one thing about how the desk runs, what would it be? Is there a ticket type or a requester that is wearing you down? Heard: __________________________________________________________________ 3. THE PERSON (always third, never first) How are you, honestly, after the last fortnight? Is there anything about your workload or hours I should know? Have you taken the breaks and the leave you are owed? If not, what is stopping you? Is there anything you need from me that you have not had? Heard: __________________________________________________________________ CLOSING "What else?" (ask it; wait) Anything you want me to raise upward, and may I name you when I do? ________ AFTER THE MEETING (the manager fills in) The one thing I will change before next time: ______________________________ By when: ________ Told the technician which one: [ ] Early-signs ticks this month, from the checklist: ____ Action if three or more: ____________ Next one-to-one: ________ (booked, not "soon")
Priyanka deleted the per-technician table from the report at twenty to seven. She replaced it with the desk's figures and a line about the arithmetic, and sent it on Monday morning instead. That is also when she sent the three e-mails she had written on Friday night. Her first one-to-one with Theo was the following Tuesday. The certificate problem came up in the second group of questions, where he described it as the best week he had had in a year. It had cost him the bottom of the table. It had been worth more than the table, and now someone had said so.
Frequently Asked Questions
I manage the desk and also work the queue most of the day. How do I find time for any of this?
My director wants per-technician figures in the monthly report. What do I say?
What if a one-to-one turns into a complaint about a colleague?
How do I leave on time when the queue is bad and the team can see it?
The staffing arithmetic shows a gap and the answer was no. Now what?
How do I notice the signs in myself as the manager?
From the QuerySolv team: we build self-hosted help desk software for Windows. QuerySolv runs as one Windows service on your own network. It takes requests by web form, e-mail, live chat, social messages and REST API. It is licensed once per installation rather than per agent per month. The free trial runs the full product for 30 days.
