The Quarterly Intelligence Review: Three Findings, No More

The last Friday of the quarter, 16:00, at Pinecrest Schools, a trust of four schools with a desk of three. Rosa has the quarter's export open, one thousand six hundred and forty tickets. Beside it is a page of notes headed "findings" with fourteen numbered lines on it. Every one of the fourteen is real. The month-end finance report, the timetable system nobody trained the teaching staff on, the September starters with no access, the catering tills, the exam-week slowness, the printers at the primary school, and eight more. The head's assistant has given her twenty minutes at Tuesday's senior leadership meeting, between the budget and the safeguarding update. Fourteen findings in twenty minutes is a list read aloud. Three is a review.

The quarterly intelligence review is a short, regular session with leadership in which the desk reports what the queue said about the organisation this quarter. It is not about the desk, which has its own review, but about everyone else. It is limited to three findings, each with a count, a cost in hours, an owner by role and an ask. And it opens with what became of last quarter's three. This article is part of our Ticket Data as Organisational Intelligence series. This is the last article, because the review is where the other five end up. The signals, the evidence and the briefs all arrive here, once a quarter, on one page.

It covers why this is a second review and not part of the first, and why a quarter. Then it covers getting from fourteen findings to three and the shape of a finding. Next come the privacy line restated for the room most likely to test it and what happened to last quarter's three. It ends with the meeting, the record over a year, what goes wrong, a short war story, and the review agenda to copy, in that order.

Two Reviews, Not One

The desk already has a quarterly review, or should. And it is about the desk: volume, backlog by age, response and resolution times, reopens, satisfaction, what the desk did and what it needs next. That review is the subject of The Quarterly Review in the reporting series, and nothing here replaces it. The intelligence review is a different page with a different question. The first asks "how is the desk doing?". The second asks "what did the desk see?". Same quarter, same tickets, different reader in the leadership team's head.

Keep them apart, on separate pages and ideally in separate slots, because a desk that mixes them makes its findings sound like excuses. "Volume is up twelve percent, partly because finance's month-end process is broken" is a true sentence that leadership hears as the desk blaming finance for its own numbers. The same finding on its own page, a week later, with the hours it costs finance, is heard as news about finance. If you run QuerySolv, the standard reports on volume, backlog by age, response and resolution times, reopen and escalation rates and satisfaction are the material for the first page. In that case, the second page is read from the same tickets with a different question. No report prints that page, because the question is about the building rather than the queue.

The second review also has a different tone. The desk's own review is the desk accounting for itself, and a little defensiveness is natural. The intelligence review is the desk being useful to people who did not ask. And the tone is that of a colleague who noticed something. The findings are offered. The owners are named by role. The asks are small. Nobody in the room is being graded, least of all the desk, which is one reason leadership comes to like this twenty minutes more than the other one.

Why a Quarter

A month is too short to confirm a pattern, because one bad fortnight makes a pile of anything. A finding reported monthly will be wrong about one time in three. That is enough to end its credibility. A year is too long, because by the time the finding is reported the owner has changed, the process has changed, or everyone has forgotten the quarter in which it mattered. A quarter is one season: long enough for a pattern to repeat, short enough for the owner to remember the tickets, and about the time an owner needs to make a small change and have it counted again. The signal tests in What the Queue Already Knows About the Company all assume two periods of evidence, and the quarter is the period.

The quarter also matches the rhythm leadership already works to. Budgets, plans and the other reviews are quarterly. A finding that arrives on that beat lands on a desk that is already in the mood to decide things. A finding that arrives on a random Wednesday lands on a desk that is in the mood to forward it. (Forwarding is how leadership says "interesting" without having to decide whether it is.)

From Fourteen to Three

Three is the number because a leadership meeting makes about three decisions in twenty minutes. A review with more findings than decisions has handed leadership a reading list. Rosa's fourteen were all real, and eleven of them had to go somewhere other than the page. Four tests do the sorting, applied in order. Is it real, by the signal-or-noise tests: it repeats, it clusters, it survives a per-head or per-user check? Is it owned: does an owner exist by role, and have they been briefed or can they be before the meeting? Is it actionable this quarter: can the ask be tried and counted before the next review? And, among the survivors, which three cost the organisation the most hours, with the owner's team's hours counted before the desk's?

Two of the seven shown are the desk's own, and they are the easiest to mistake for findings, because they are real patterns with real hours. They do not belong in this review; they belong in the desk's own review as things the desk is doing. The intelligence review is about the organisation. And a desk that fills a slot with its own printer problem has spent a decision on itself. The watch list takes the rest, one line each, in an appendix nobody has to read and everybody can. A review with fourteen findings has no findings. It has a spreadsheet with ambitions.

Choosing by the owner's hours rather than the desk's is the rule that most changes the selection. The printer problem costs the desk more than the finance report does. And the finance report goes on the page. That is because the review is read by people who run the organisation. They weigh an hour of a finance analyst's time differently from an hour of a technician's. That is not a judgement about whose time matters. It is a judgement about which findings a leadership meeting can act on, and leadership can move a report; it cannot fix a printer.

The Shape of a Finding

A finding is one short paragraph with five parts, read in the same order every quarter so that the room learns the shape. The pattern, as a count per run or per quarter, with the team and the system. The cost in hours, the owner's team first, then the desk, as ranges. The owner, by role. The one ask. And the status: briefed, agreed, in progress, done, or declined, with the date. Nothing else. No history of how the desk found it, no theory of why it happens, no example tickets read aloud, which are in the brief the owner already has.

Here is a finding written badly, from Rosa's first draft: "Finance keeps breaking the month-end report and it's a nightmare for the desk every month." Here is the same finding in shape: "Finding one, the month-end report. Twenty-two tickets this quarter, on the last two working days of each month, from the finance team, because the report runs during the overnight batch. Roughly eight hours of finance analysts' time a month and two of ours. Owner: the finance manager. Ask: run the report on the first working day instead, for one month. Status: briefed on the twelfth, agreed, first count due after the next month-end." Sixty seconds to read, and every question the room could ask is already answered except "did it work?", which is next quarter's first item.

The status line is the part that makes the review a record rather than a presentation. A finding that is "briefed, agreed" is on its way. A finding that is "briefed, declined" is in the room with the owner's peers, neutrally, with its hours beside it. That is all the pressure the desk ever applies and usually all it needs. The craft of the briefing that produced the status, and what the desk does when the owner says no, is the previous article in this series, Taking the Signal to the Owner Without Making an Enemy. The review inherits its statuses and does not redo its conversations.

The Privacy Line Restated

The review is the moment the privacy line is most tested. That is because the room is senior, the findings are interesting, and the most natural question a senior person can ask about a pattern is "who?". Which team is the worst. Which manager is sitting on the approvals. Which person raises all these tickets. The questions are not malicious; they are how leadership is used to thinking. The desk's answer is the same every time and is written on the agenda so that nobody has to improvise it. This review reads patterns and never people. It names processes, systems and teams. Where a team is small enough that naming it names a person, it names the process instead. Per-person counts are not in the review and are not available on request.

Say why, once, in one sentence, because leadership deserves the reason: the queue only works while people trust it, and a review that names a person ends the data for everyone. Every finding on the page exists because somebody wrote down, honestly, what stopped their work. The first time a name is read out in this room, that honesty is over. And the following quarter's page will be shorter and quieter and will stay that way. The desk is not protecting individuals from leadership. It is protecting leadership's only source.

The same line covers the desk's own people. A finding never says which technician handled the pile, and the review never carries a per-technician figure of any kind. Those exist only inside the coaching rule that Dashboards That Motivate Instead of Surveil sets out, read with the person and never ranked. A leadership team that is told this once, plainly, in the first review, almost never asks again. The ones that keep asking are telling the desk something about the organisation that no ticket could.

What Happened to Last Quarter's Three

The review opens with last quarter's three, before this quarter's, because results are what earn the desk the next twenty minutes. For each: the finding in one line, the ask, the status now, the count before and the count after, and the hours returned to the owner's team. "The month-end report: moved to the first working day; twenty-two tickets the quarter before, four this quarter, about seven hours a month back to finance." Four lines, and the room has learned that a decision it took last time did something. That is rarer in leadership meetings than it should be.

Credit goes to the team and its manager by role, loudly, and never as a name in the written record unless the manager asks for it. "Finance moved the report and the tickets fell" is the sentence; the finance manager is in the room and knows who moved it. A finding that was declined is reported in the same neutral line as one that succeeded, with the count still running. For example: "the records system slowness: owner briefed, declined to raise with the vendor; thirty-one tickets this quarter, roughly the same as last; about twelve hours of office staff time." No comment follows it. The room will supply its own, and the owner's peers supply it more effectively than the desk could.

A finding that did not move is "no change yet", not an accusation. Changes take a run or two to show. Owners get busy. The desk's job is to report the count on the date and not to editorialise about the gap. A finding reported as "no change yet" for two quarters running becomes "agreed, not implemented". That is a status, not a verdict, and one owners prefer not to see beside their role twice. The count does the work. The desk reads it out.

The Meeting

The meeting takes twenty minutes, in this order. Start with one sentence on what the review is and is not. Then come last quarter's three, with results, and this quarter's three, in shape. Next comes the watch list in one breath. Then comes one question to the room: "Which of these do you want us to pursue, and is there anything you would like counted next quarter?" The question is the point of the meeting. It turns the review from a report leadership receives into a brief leadership gives. And a leadership team that has asked the desk to count something will read the answer. Leave the page behind. Send it afterwards. Do not send the export.

Resist the urge to defend a finding that is questioned. "Are you sure it's the batch and not the network?" is answered with "the owner's team reproduced it with the batch stopped; I can send the detail", and the meeting moves on. A finding that cannot survive one sceptical question in twenty seconds was not ready for the page. And the right response is to say so and bring it back next quarter with the evidence. That costs the desk nothing and buys it a reputation for not overclaiming. The desk that argues in this room wins the argument and loses the slot.

The bad quarter, where the desk's own numbers are poor, is still a quarter in which the queue saw things. The intelligence review should go ahead on its own page as usual. A desk under pressure is tempted to use its findings as cover, which is exactly the mixing the first section warns against. The separation protects the findings from the desk's bad month as much as it protects the desk's review from sounding like excuses.

The Record Over a Year

Keep one page per quarter, unchanged after the meeting, with the statuses updated the following quarter on the new page rather than on the old. After four quarters the set of pages is a short history of what the organisation learned about itself from its own tickets. It is also a history of what it chose to do about each thing. That is a document no other team in the building has and which the desk did not set out to write. Read it before each review. The findings that recur across quarters, briefed and declined and counted again, are the organisation's standing problems. The page that lists them beside their hours per year is the single most persuasive thing the desk will ever hand upward.

The record also feeds the desk's other reviews. The hotspot review, in The Quarterly Hotspot Review, is the desk's own quarterly cut of devices, applications and processes. And it is where most intelligence findings are first noticed. The two reviews share an export and a spreadsheet and differ in who reads them. Pull the quarter's tickets once, through whatever export or interface the tool offers, cut them once, and write both pages from the same sheet. A desk that runs two exports for two reviews will shortly have two sets of numbers, and will be asked which one is right.

What Goes Wrong

Four things go wrong, and each has been described above by its absence. The review becomes a complaints session, because the findings were written in the desk's voice about the desk's pain rather than in the owner's voice about the owner's hours. The five tone rules in the previous article fix it. The desk brings fourteen, because every finding felt too real to leave out. The four tests and the watch list fix it. Leadership asks for names and gets them, because the desk had not prepared the sentence. The agenda's standing line fixes it. If it fails once it will not fail twice, because the following quarter's page will have shown the cost. And nobody does the follow-up count, because the date was never written down. The status line with its date fixes it. The review's opening item makes the missing count impossible to hide.

There is a fifth, quieter failure, which is that the review works so well that leadership starts to see the desk as the organisation's inspector. In that failure, the desk starts to enjoy it. The cure is the tone: offered, never demanded; patterns, never people; three findings, no more. A desk that reads the organisation and remains a colleague gets to keep doing it. A desk that reads the organisation and becomes its auditor is read, in turn, by everyone else, and finds that the diary has stopped being written.

A Short War Story: The Eleven Findings at Oakridge Dental

Oakridge Dental, a group of six practices with a desk of one and a half, had its first intelligence review when Ana, who ran the desk, was given a slot at the partners' meeting. She brought eleven findings, each real, each with a count, on four pages. The meeting ran forty minutes over. The partners asked which receptionist was raising all the appointment tickets. Ana declined to say. The review ended with no decisions and a general feeling that IT had been complaining. The appointment system's double bookings were the largest finding on the page. They were caused by a reception process that let two practices book the same hygienist. They cost about ninety tickets and roughly as many patient apologies in the following quarter while nothing was done. Ana's second review had three findings on one page. It opened with the privacy line read aloud. It put the double-booking finding first with the hours of reception time beside it and one ask. The ask was a shared booking rule that the practice manager could set in an afternoon. The partners agreed in the room. The double bookings fell to a handful the following quarter. The review was given a standing slot, which it has kept. Three findings got three decisions; eleven had got none. The quarter in between had cost ninety tickets and a room full of people who thought the desk was there to complain.

The Quarterly Intelligence Review Agenda

The agenda below is the page the desk brings and the page it leaves behind. It fits on one side. It is read in the order shown. It carries the privacy line as a standing item so that the answer to "who?" is already on the table before anyone asks. Fill the three findings in shape, bring last quarter's three with their counts, list the watch list in one line each, and end with the question. Keep each quarter's page unchanged after the meeting; the next quarter's page carries the updates.

QUARTERLY INTELLIGENCE REVIEW — AGENDA, TWENTY MINUTES
Quarter: ____________   Desk: ______________   Presented by: ______________
Attending (roles): ________________________________________________

0. WHAT THIS IS (one sentence, read aloud)
   What the help desk queue said about the organisation this quarter: patterns,
   never people; processes, systems and teams, never a person; the desk's own
   performance is on a separate page.

1. LAST QUARTER'S THREE (eight minutes)
   Finding ____________________  Ask ________________  Status ______________
      Before ______  After ______  Hours returned to the team ______
   Finding ____________________  Ask ________________  Status ______________
      Before ______  After ______  Hours returned to the team ______
   Finding ____________________  Ask ________________  Status ______________
      Before ______  After ______  Hours returned to the team ______

2. THIS QUARTER'S THREE (eight minutes; each in shape)
   1. Pattern and count per run: _______________________________________
      Cost: their team ____ to ____ hours; the desk ____ to ____ hours
      Owner (role): ______________   Ask: ______________________________
      Status and date: ______________
   2. Pattern and count per run: _______________________________________
      Cost: their team ____ to ____ hours; the desk ____ to ____ hours
      Owner (role): ______________   Ask: ______________________________
      Status and date: ______________
   3. Pattern and count per run: _______________________________________
      Cost: their team ____ to ____ hours; the desk ____ to ____ hours
      Owner (role): ______________   Ask: ______________________________
      Status and date: ______________

3. THE WATCH LIST (one minute; one line each; no discussion)
   ____________________________   ____________________________
   ____________________________   ____________________________

4. THE QUESTION (three minutes)
   Which of these do you want pursued?  Anything you would like counted next quarter?

5. NOT IN THIS REVIEW
   The desk's own numbers (separate page). Names. Per-person counts, for requesters
   or for technicians. Theories not yet counted.

Rosa's twenty minutes on Tuesday ran to eighteen. The month-end report was agreed in the room. The deputy head asked for the timetable guides by half term. The head asked HR to bring the starter finding to the next meeting with a plan. The watch list was read in one breath and nobody asked about it. Then the head asked whether the desk could count how many tickets came from the two schools that shared a site. Rosa had not thought to make that cut. It turned out, the following quarter, to be finding one.

Frequently Asked Questions