Home › Topics › Experience Over Metrics › Reporting It

Reporting Experience to Leadership Beside the SLA Table

Tuesday, 09:05, at Greyfriars Academy. Elin runs a desk of three for a school with a hundred and ninety staff. She is in the monthly operations meeting with her report on the table. It is one page, the service-level table at the top, every row green for the fourth month. Gareth, the head of operations, reads it and puts it down. He asks the question he has asked in some form every month this term: "Then why do the teachers complain to me?" Elin knows the answer. She has the ten-ticket reading in her bag, the effort answers in a spreadsheet, and three comments she could quote from memory. None of it is on the page, because the page has a table for time and nowhere to put a feeling. A feeling is what Gareth is being told about in the staff room.

Reporting experience to leadership means adding a small, fixed panel to the monthly report, beside the service-level table and never instead of it. It has two numbers, one quote and one moment story, chosen by rules written in advance. It has a line saying what the panel does not claim. The format of the monthly report itself, the one page a director will actually read, is the subject of The Monthly Report They Will Read. This article does not redesign it; it adds a panel to it. This article is part of our Experience Over Metrics series, and it is the last one. Reporting is where the other five either reach leadership or stay in the desk lead's bag.

It covers why the service-level table needs a neighbour, and the two numbers and how they are stated. Then come the one quote and the rule for choosing it, the moment story, and placing the panel beside the table. Next come what not to claim, what to do when the experience numbers are bad, and the quarterly version. It ends with a short war story and the experience report line to copy, in that order.

Why the Service-Level Table Needs a Neighbour

Leadership hears about the desk in two ways, and the report only carries one of them. The first is the report, which says how fast tickets were answered and resolved against the targets of the service level agreement, the SLA. Its table is green when the desk is keeping its time promises. The second is the corridor, the staff room and the car park, where people say how getting help felt. That is a different thing and is said in different words: "they never tell you anything", "I had to ask three times", "it got closed on me". A director who reads a green table every month and hears the corridor every week does not conclude that the corridor is wrong. They conclude that the table is measuring something else, and they are right.

The panel exists to put the corridor on the page, in a form the desk controls and can defend. If the report carries the thing leadership is hearing elsewhere, the report becomes the place the conversation happens, with counts beside it. If it does not, the conversation happens without the desk, in the words of whoever complained last, and the green table becomes evidence that the desk is not listening. Gareth's question is not hostile. It is a request for a row that is not there.

The neighbour is small on purpose. One page is the limit of what gets read, and the service-level table has earned its place on it. The experience panel takes a quarter of the page, is the same size every month, and never grows into a second page of charts. A panel that is the same shape every month is read in the same glance every month. The glance is where the pattern shows. The series on reporting has already made the case for the numbers to show and the numbers to keep, in The Numbers to Show and the Numbers to Keep. This panel shows two and keeps the rest.

The Two Numbers

The first number comes from people and the second from tickets, and that is why there are two. The first is the share of requesters who answered the survey and said they got what they needed, stated as counts: "31 of 38 who answered". It is the satisfaction figure the desk may already report, moved beside its neighbour and stripped of its decimal place. The second is the share of resolved tickets on which the requester had to chase, stated the same way: "27 of 164 tickets". It is the effort figure that covers everyone, because it is counted from the ticket history rather than from whoever clicked. It is the one that answers Gareth's question most directly. The method behind both, and the effort question that rides beside the satisfaction one, is in Measuring Experience Honestly. The panel reports what that article measures.

Counts, not percentages, for two reasons. A count carries its own sample size, so "31 of 38" tells the reader at once that this is a small, self-selected group. "82 percent" invites them to believe something about all hundred and ninety staff. And a count is the same shape month after month. So a reader who saw "27 of 164" last month and sees "19 of 171" this month has read the trend without a chart. If a director asks for the percentage, give it, in speech, with the counts still on the page. The page is the record; the conversation can be rounded.

Each number carries one line of context underneath, in the smallest type on the page. For the first, it says how many tickets were resolved, so the reader can see how many did not answer. For the second, it says what counts as a chase, in ten words, so the figure cannot be misread as complaints. "A chase is a message asking for status before any update" is enough. The context lines never change wording, because a definition that moves is a number that cannot be compared. If you run QuerySolv, the standard reports already carry the service-level half, response and resolution times and satisfaction. In that case, the panel's first number is a figure you have and its second is a filter you run. The writing is the only new work.

The One Quote

The quote is one requester comment, verbatim, anonymised, chosen by a rule that is written on the page beside it. It is the most-read thing on the report, because it is the only sentence on it that a person said. That is exactly why it needs a rule. A quote chosen by whoever makes the report will be the kindest one in a good month and the most defensible one in a bad month. Leadership will work that out within a quarter. After that, every quote reads as a selection, which is worse than no quote at all.

The rule that works is the comment that best describes the moment most comments pointed at this month, whether or not it is kind. If eleven of the month's comments mention silence while waiting, the quote is the clearest of those eleven. If the clearest one is "I didn't hear anything for four days and had to ask twice", that is the quote. If the month's comments mostly mention being kept informed, the quote is the clearest of those. In that case, it will be a kind one, and it will be believed, because last month's was not. The rule is printed under the quote in a few words: "chosen as the comment closest to the month's most common theme". A rule the reader can see is a rule the reader can trust, and a rule they cannot see is a choice.

Anonymised means no name and no department small enough to identify a person. "A teacher" is fine where there are ninety teachers, and "the deputy head" is not. The requester is never told their comment was quoted, and the comment is never used against the technician it describes. That is the condition under which comments stay honest. (The quote is for leadership. The technician gets the conversation, not the slide.)

The Moment Story

The moment story is one ticket from the month, told in four sentences from the requester's side. It tells what they asked, what happened at the moment that decided how it felt, and what the desk changed because of it. It also tells what the count has done since. It is the panel's only narrative. It does what the numbers cannot, which is to show leadership what a chase or a silent close looks like to the person inside it. A director who has read one moment story knows what "27 of 164" means in a way no definition line can give them.

The story is chosen from the month's ten-ticket reading, by rule again. It is the ticket that best shows the moment the desk worked on this month. In a month where the desk built its update cadence, the story is a ticket where the update arrived before the chase and the requester said so. In a month where the reading found closes without a question, the story is one of those, told honestly, with what changed. The story alternates between a ticket that went well and one that did not over the year, roughly, and the rule says so. A panel that carries only good stories is a newsletter, and a panel that carries only bad ones is a confession. Neither is a report.

Four sentences is a limit, not a target. "A teacher raised a ticket about a projector on a Monday and heard nothing until Thursday, when it was closed as fixed; she had asked twice in between. The silence was the moment, and the reading found four tickets like it. We now write to every open ticket at least every two working days with the next time, and the morning starts with those. Chases fell from 27 to 19 the following month." Nothing in that is a claim the ticket history cannot support, and the fourth sentence is the only one with a number in it. Where the desk's event history holds every message with its time, the story is read straight from the ticket. In that case, it is not reconstructed from memory, and it is not improved.

Placing It Beside the Table

The panel sits to the right of the service-level table, or directly under it on a narrow page. It is the same width as the table, in the same type, with the same month in its heading. The two are read in one glance, which is the design. A reader whose eye moves from a green row to "27 of 164 had to chase" has seen the whole truth about the month without turning a page. A reader who sees the green row alone has seen half of it. Nothing separates the two except a rule and a heading. They are one report about one desk, measured two ways.

No gauge, no faces, no colour fill behind the numbers. The service-level table earns its green by comparing a count to a target the desk agreed. The experience numbers should earn theirs the same way, against the targets set in Experience Targets Beside Service Levels. The targets are printed in a column beside the count: "27 of 164; target fewer than 1 in 8 by year end". A number with its target is a measurement. A number with a smiling face is a mood, and the mood has a way of being the report-writer's rather than the requester's.

The panel also keeps its place when the month is uneventful, which most months are. A panel that appears in bad months and vanishes in good ones is read as an alarm rather than a measurement. Leadership will come to associate its presence with trouble. Same panel, same shape, every month, including the months when the quote is kind and the story is short and nobody has anything to add. Those months are the ones that build the trust the bad months will spend.

Remember: the experience panel is two counts, one quote and one four-sentence story, chosen by rules printed on the page. It is placed beside the service-level table every month in the same shape, with a line saying what it does not claim.

What Not to Claim

The panel claims what the tickets and the survey can show, and nothing beyond that, and the line saying so is printed on it. It does not claim that staff are more productive, that anyone stayed in their job, that hours were saved across the organisation, or that the school is better run. The desk has no evidence for any of those and a director who asks for it will find none. It does not claim that "users love IT", because thirty-eight people answered. It does not compare the desk's figures to any industry number, because the counting rules differ and the comparison is a sentence that cannot be checked. The reputation the desk earns from reporting honestly is the subject of the series on IT's internal brand. The series makes the same point from the other side. The brand is built on claims that survive questioning, and lost on the first that does not.

It also does not claim a trend from one month, or a cause from one change. "Chases fell from 27 to 19" is a fact. "The new cadence cut chases by a third" is a hypothesis that one month cannot confirm. The honest form is "chases fell from 27 to 19 in the month after we started the cadence. We will know by the quarter whether it holds". A desk that claims causes from single months will be right often enough to feel confident and wrong often enough to be caught. Being caught once costs more than a year of careful wording saves.

And it never carries a per-technician figure. The experience numbers by person exist for coaching, from a quarter's sample, in a conversation between the technician and the desk lead. They do not go on a page that leadership reads. A name beside a number on that page is a ranking whatever the heading says. A desk whose technicians know their names might appear there has a survey its technicians are managing. The panel describes the desk. The desk is the unit that leadership can act on.

When the Experience Numbers Are Bad

When the numbers are bad, lead with them. The panel's first line in a bad month is the bad count, followed by the cause as far as the reading found it. Next come the change the desk is making, and the month in which the change should show. "Chases rose from 19 to 41 of 158. The reading found most of the rise in tickets handed to the new technician in his first month, where the introduction was not yet habit. He is paired with Mara for handoffs until the end of next month; we expect the count back under 25 by then." Four sentences, no defence, no adjective, and a date by which the report can be checked. The general method for a bad month, and what leadership does with a desk that reports one honestly, is in When the Numbers Are Bad. The experience panel follows it exactly.

A bad month in the experience panel beside a green service-level table is the most useful report the desk will ever produce. It shows leadership the thing they were hearing in the corridor, with a number on it and a plan under it, before they had to ask. Gareth's question stops being a question the month the panel answers it first. The green table did not lose credibility that month. It gained a neighbour that told the truth, which made the green more believable, not less.

Do not soften a bad month with a kind quote. The quote rule stands in bad months as in good: the comment closest to the most common theme. In a bad month the theme is the bad thing. A panel with "41 of 158 had to chase" beside a quote saying "brilliant, thanks" has been edited, and will be read as edited. The quote that matches the count is the one that makes the plan under it credible.

The Quarterly Version

The quarterly version is the same panel with three months side by side and the targets printed against them. It adds two lines the monthly cannot carry, in two sentences. They cover what the desk changed this quarter and what the counts did afterwards. They also cover what the listening beyond the survey turned up, the department conversation and the silent-user check. The quarterly is where a trend can be claimed, cautiously, because three months is the smallest run in which one month's weather stops dominating. It is where the experience targets are reviewed, met or not, with the levels for the next quarter stated. The quarterly review as a whole, and how the desk's numbers sit inside it, is a subject the reporting series owns. The experience panel contributes a quarter of a page to it.

The quarterly is also where the moment story is replaced by the year's pattern. Instead of one ticket, the quarterly says which of the five moments the reading found weakest each month, in one line. That lets leadership see the desk working through them: "waiting update in month one, handoff in month two, close in month three". It is a sentence that shows a desk improving by design rather than by luck. It is the single most persuasive line Elin has ever put in front of Gareth. It told him what the desk was going to fix before it had fixed it, and then it did.

A Short War Story: The Cheerful Quote at Calder Transit

Dmitri ran the desk at Calder Transit, a bus operator with four depots. For eight months his monthly report carried a satisfaction figure, a verbatim comment, and nothing about effort. The comment was chosen each month by Dmitri, and each month it was the kindest one the survey had produced. The report was going to the operations director and Dmitri wanted the desk to look good. The director heard the depots at a staff forum in the spring. A shift supervisor said that raising a ticket was "like posting a letter into a wall". The director asked Dmitri why eight months of reports had quoted eight people who disagreed. The cost was not the forum. It was the headcount case Dmitri presented two months later, built on honest volume figures. The director declined it with the words "your numbers are selective". The cost was also a year in which every figure from the desk was checked by someone else before it was believed. Dmitri rebuilt the report with a chase count from the tickets, a quote chosen by a printed rule, and a line saying what the panel did not claim. It took three quarters for the director to stop asking who had chosen the quote. The lesson Dmitri gave his successor was written on the inside cover of the report template. A quote chosen to please is a number the reader will never trust again. The number you most need them to trust will be the one you present next.

The Experience Report Line

The panel below is the quarter of a page that goes beside the service-level table, blank, with the rules printed where the reader will see them. The two counts come from the monthly measurement, and the quote and the story come from the ten-ticket reading. The "not claimed" line is fixed wording that does not change month to month. Fill it in the same afternoon the service-level table is filled in, by the same person, and keep the shape identical every month. The quarterly block at the bottom is used four times a year, with the three monthly counts carried forward and the targets beside them. Nothing on the panel is a joke, a gauge or a face, and nothing on it has a decimal point.

EXPERIENCE THIS MONTH — BESIDE THE SERVICE-LEVEL TABLE
Desk: ______________________   Month: ____________   Resolved tickets: ______

THE TWO NUMBERS                                          Target
  Got what they needed:   ____ of ____ who answered      ____ in ____ by ____
    (____ of ____ resolved tickets answered the survey)
  Had to chase:           ____ of ____ resolved tickets  fewer than ____ in ____ by ____
    (a chase is a requester message asking for status before any update from the desk)

THE ONE QUOTE (verbatim, anonymised)
  "____________________________________________________________________"
  Chosen as the comment closest to this month's most common theme: ______________________

THE MOMENT STORY (four sentences, from the requester's side)
  What they asked:            ________________________________________________
  The moment that decided it: ________________________________________________
  What we changed:            ________________________________________________
  What the count did since:   ________________________________________________

WHAT THIS PANEL DOES NOT CLAIM
  Nothing about productivity, retention, hours saved or money; no comparison with any outside
  figure; no trend from a single month; no cause from a single change; no figure by technician.

IF THE NUMBERS ARE BAD (this block replaces the story in a bad month)
  The count: ____________   The cause, as far as the reading found it: ______________________
  The change we are making: ______________________   Expected to show by: ____________

QUARTERLY BLOCK (every third month)
                         Month 1      Month 2      Month 3      Target      Met?
  Got what they needed   ____/____    ____/____    ____/____    ________    yes / no
  Had to chase           ____/____    ____/____    ____/____    ________    yes / no
  Weakest moment found   __________   __________   __________
  What we changed this quarter and what the counts did afterwards:
  ____________________________________________________________________
  What the listening beyond the survey turned up (department conversation, silent-user check):
  ____________________________________________________________________
  Levels for next quarter: ______________________

Elin's next report had the panel on it: 31 of 38 got what they needed, and 27 of 164 had to chase. It also had a quote about four days of silence, and a story about a projector. Gareth read the table, read the panel, and asked a different question for the first time all term: "What are you doing about the 27?" She told him. It was on the page. The month after, the number was 19, and the question after that was "will it hold?", which is the question a report is for.

Frequently Asked Questions

Why counts rather than percentages?
A count carries its own sample size, so "31 of 38" tells the reader at once that a small, self-selected group answered. A percentage invites them to believe something about everyone. Counts also keep the same shape month to month, so a trend is visible without a chart. Give the percentage in conversation if asked, with the counts still on the page.
Should the quote ever be a critical one?
Yes, whenever the rule says so. The quote is the comment closest to the month's most common theme. In a month where most comments mention silence or chasing, the quote describes silence or chasing. A panel that carries only kind quotes is read as a selection within a quarter, after which no quote on it is believed. The printed rule is what lets a kind quote be trusted when the month earns one.
What if we have too few survey answers for the first number?
Report the counts anyway, with the sample line underneath, and let the second number carry the weight. The chase count covers every resolved ticket whether or not anyone answered. On a very small desk, report the first number quarterly rather than monthly and say so on the panel. Never fill the gap with a percentage from a dozen answers.
Can we show experience by department or by technician?
By department, yes, in the quarterly block, when a department's counts are large enough to mean something. The purpose must be to find where the desk should look next. By technician, never on a page leadership reads. Those figures exist for coaching, from a quarter's sample, in a private conversation. A name beside a number on a leadership report is a ranking whatever the heading says.
How do we handle a month where nothing much happened?
Fill the panel exactly as usual, with the counts, a quote chosen by the rule, and a short story, even if the story is a ticket that went well. The panel's value comes from appearing in the same shape every month, so that it is read as a measurement rather than an alarm. Quiet months are the ones that build the trust a bad month will need.
Leadership wants to know what the experience work is worth in money. What do we say?
Say plainly that the panel does not claim money, productivity or retention, because the desk cannot evidence them. Say that the honest effects it can show are fewer chases to answer, fewer reopens and fewer requesters going around the desk. Where a cost figure is genuinely wanted, build it in hours from the chase and reopen counts. Use the worked-example method in the library's economics and ROI series, as a range, and keep it off the experience panel.

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.