Presenting Support Economics Without a Spreadsheet War

Thursday, 14:00, at Redbrook Engineering, in the small meeting room with the projector that takes four minutes to warm up. Priya, who runs the four-person desk, has a quarter of work behind her: the cost map, the cost-per-ticket read, the waiting count, the payback sheets, the habit table. It is all in a workbook with nine tabs. Across the table, Owen from finance has his own workbook, built from her first cost map and his hourly rate. The operations director, Ruth, has twenty-five minutes. Priya opens tab one. Owen asks which hourly rate she used. She says she did not use one. He says his model does, and names it, and Priya says it looks high, and Ruth looks at the clock. Nineteen minutes later the two of them have agreed on nothing about the rate. Ruth has asked them to "come back when the numbers line up". The desk's quarter of work has produced a scheduling conflict.

A spreadsheet war is a meeting in which two people have brought two models of the same thing built on different assumptions. The time is spent on the assumptions instead of the decision. It is the commonest way for a desk's economics to fail, and the desk nearly always loses it. The other side owns the assumptions that are about money and the desk owns only the ones that are about tickets. This article is about not having the war. Present what support costs in hours, in ranges, on one page, around one number you can defend. That way, the meeting ends in a decision rather than a rematch. This article is part of our The Economics of Support series, and every number in it is a worked example from a fictional desk.

It starts with what a spreadsheet war is and why the desk loses it. Then come hours as the currency and handing over the conversion, and ranges and how to say one aloud. Next are the one page and what goes where, the number you will defend, and what to concede before you are asked. Then come the three questions you will get and presenting the bad numbers. A worked example of a fictional desk's page follows. A short war story and the economics one-pager to copy finish the article.

Why the Desk Loses the Spreadsheet War

The desk loses because it has brought the wrong units into a room where the other side owns those units. The moment a support figure is expressed in currency, it depends on an hourly rate. The rate is a finance assumption: loaded or unloaded, with or without overheads, this year's or the budget's. Finance knows those rates and the desk does not, so every currency figure the desk presents is a figure finance can correct. Correcting it is what the meeting becomes. Priya's nine tabs had no rate in them, which was right. Owen's had one, which was his job. The war started anyway because the two workbooks were side by side and the rate was the only visible difference between them.

The desk also loses because a workbook invites a workbook. When one side arrives with nine tabs, the other side builds something comparable, and the decision-maker is asked to referee rather than decide. A single page with five lines on it invites a question, not a counter-model. Nobody builds a workbook to argue with a page. The one-page form is not a simplification of the economics; it is a choice about what kind of meeting to have. The series on reporting upward, Reporting to Leadership, is where the form itself is taught. This article borrows it and fills it with costs.

The last reason is that the desk arrives wanting to be believed about everything. A presenter who defends every figure equally will be argued with about the weakest one. A quarter of counting produces solid figures, estimated figures and one-week tallies, and they are not equally defensible. If the presenter cannot say which are which, the room will find out by testing them. It will start with the softest, which is always the requester waiting line. Lead with the solid number and label the soft ones as soft, and there is nothing to attack. (People do try anyway. It is a meeting.)

Hours Are the Currency: Hand Over the Conversion

Present every figure in technician hours or requester hours and never in money, and say at the start that the conversion is finance's to do. This one rule prevents most spreadsheet wars, because it hands the one assumption the desk cannot defend to the person who can, in front of the decision-maker. "Everything on this page is in hours. Owen, you have the rates; if Ruth wants a currency figure, you can give her a better one than I can." Owen is now an ally with a job rather than an opponent with a model. The only thing left to argue about is the hours, which came from tickets and tally sheets that Priya can show.

Hours are also the unit the decision-maker can picture. "A hundred and sixty hours a month of the desk's time goes on rework, walk-ups and re-typing" is a sentence Ruth can turn into a person: that is most of one technician, every month, doing things twice. The case for hours as the honest currency is made at the start of this series, in Where the Money Actually Goes. The presenting rule is simpler: hours on the page, rates in someone else's hands, and say so out loud.

Keep requester hours and desk hours as two lines with two labels, and resist the temptation to total them for effect in the room. A presenter who says "support costs the business fifteen hundred hours a month" has just invited the question "whose hours?", and the honest answer sounds like a retreat. Say the two lines separately from the start and there is no retreat to make.

Ranges, Not Points, and How to Say One Aloud

Every estimated figure on the page is a range, and the presenter says the range as a range rather than apologising for it. "Requesters wait somewhere between five hundred and a thousand hours a month" is not a vague statement. It is a precise statement about how much is known, and a finance reader hears it as such, because finance deals in ranges all day. A point estimate, "seven hundred and sixty hours", sounds more confident and is less so. The first person to ask how it was derived will conclude, correctly, that the point was a range dressed up. Say the range. Then say, in one sentence, where it came from.

Ranges also take the fight out of the room, because there is nothing to correct. A point can be wrong; a range can only be too wide or too narrow, and a presenter who has chosen the width from the sample size can say so. The presenter who argues for a point is defending a position; the presenter who offers a range is describing the evidence, and people argue with positions, not with descriptions.

Counted figures are stated as counts with their source, not widened into ranges for the sake of consistency. "About four hundred and seventy ticket hours, from the ticket records and a week of tally sheets" is a figure with a method rather than a range with a hedge. The page mixes the two kinds and labels them. A reader who sees both labels on one page knows the presenter knows the difference, which is most of what credibility is.

The One Page and What Goes Where

The page has a question at the top, five lines in the middle, one request or one observation at the bottom, and a method note in small type, and nothing else. The question is the one the decision-maker is actually asking, in their words, such as "What does support cost Redbrook, and where could it cost less?" The five lines are the five figures this series has built. They are the desk's hours and the removable share of them, the requesters' waiting range, and the cost per ticket with its reading. Then come the habits priced with their fixes and the investments in hand with their payback. The bottom is what the presenter wants to happen next. The method note is one line per figure saying where it came from. The format itself is taught in The Monthly Report They Will Actually Read. The economics page follows it exactly and changes only the content.

Page two is the workings, and it exists so that page one does not have to defend itself. Each of the five lines on page one has a paragraph on page two with the source, the sample, the range's derivation and the oddities of the month. The presenter brings page two but does not present it. When a question is asked, the answer is on page two and can be read aloud; when none is asked, page two stays in the folder. If you run QuerySolv, the standard reports on volume, backlog by age, resolution times and reopens are where most of page two comes from. It comes as a printout rather than a model. In any tool, the point is that the workings are a reference, not a presentation.

Translate every line into the organisation's own terms before it reaches the page. "Rework" becomes "tickets we fixed twice", and "overhead" becomes "re-typing and meetings about the queue". "Blocked requester" becomes "a person who could not work until we got to them". A decision-maker who has to translate a line while reading it will skip it. The craft of that translation is the subject of Translating Tickets Into Business Language. The economics page is where it earns its keep. The cost lines are the ones most likely to be acted on and least likely to be understood in the desk's words.

The Number You Will Defend

Choose one figure on the page that you can rebuild from scratch on a whiteboard, know its method cold, and let that figure be the one the meeting turns on. The rest are context. At Redbrook the figure was the removable share of the desk's own hours: about a hundred and sixty hours a month on rework, walk-ups and overhead. It came from the cost map, built from the ticket list and a tally week by four named people. Priya could produce every piece of it. It is a counted figure, it is the desk's own time, and it needs no rate and no sample of requesters. It leads directly to the request at the bottom of the page: permission to spend some desk hours removing it. A number like that cannot be argued with, only asked about, and every answer makes it stronger.

The number you will defend is chosen for its defensibility, not its size, which is why it is almost never the waiting line. The waiting line is the biggest figure on the page and the softest. A presenter who makes it the centrepiece has chosen to hold the meeting on the one figure the other side can reasonably doubt. Put it on the page, label it as an estimate with its sample, let it do its own quiet work on the reader, and defend the counted figure instead. The soft figure persuades; the hard figure survives. Build the meeting on the second and let the first travel home with people.

Rehearse the rebuild, out loud, once, because being able to do it is the whole point. "Six hundred hours from four people's contracts minus leave. Three hundred and ten on handling, from timestamp samples of the six biggest ticket types times their counts. Fifty-five rework, from the reopened and reassigned tickets. Sixty walk-ups and forty-five overhead, from the tally week. The last two plus the rework is a hundred and sixty." Forty seconds, no spreadsheet, every step checkable. A presenter who can do that in the room has ended the spreadsheet war before it starts. There is no spreadsheet to war with, only a whiteboard with five lines on it.

What to Concede Before You Are Asked

Say the weaknesses first, in one sentence each, before anyone finds them. "The waiting figure is an estimate from forty tickets, which is why it is a range. The walk-up and overhead figures come from one week of tally sheets, so they could be a fifth out either way. The walk-ups were not logged before this quarter, so last year's ticket counts are low." Each concession costs nothing, because each is true and was going to be discovered anyway. The room cannot attack a weakness the presenter has already named; it can only agree, and agreement is a habit that carries into the rest of the meeting.

Concede the direction of your own interest too, because everybody in the room knows it. "I run the desk, so I would say this. The figures are from the ticket system and the tally sheets, and Owen is welcome to check any of them." The second sentence is the one that matters, and it only lands if the first has been said. A presenter who pretends to be a neutral analyst of their own department is not believed; one who says plainly that they have an interest and then shows the sources is. The difference is a sentence, and it is the sentence most desk leads skip because it feels like weakness. It is the opposite.

Do not concede what is solid, and know the difference. The desk's own hours, the ticket counts, the reopen count and the approval log are counted. A presenter who hedges them to seem modest has thrown away the hard figures to look even-handed about the soft ones. Concede what is estimated. State what is counted. The anticipating of what will be pushed back on, and the preparing of the answer, is the subject of Anticipating the Pushback. The economics page's version is simply to say the pushback yourself before anyone else does.

The Three Questions You Will Be Asked

The first question is "how do you know?", about whichever figure caught the reader's eye. The answer is the page-two paragraph for that line, read aloud in two sentences. "From a random sample of forty tickets last month, each classified as blocked, degraded or merely waiting, in working hours. Twelve were blocked at three and a half to six hours each, and so on; the sheet is here." That is the whole answer. A presenter who answers "how do you know" with a method has answered; one who answers with a defence has started a war, and the question was not an attack.

The second question is "what would it cost to fix?" The answer is in hours, from the payback sheets and the habit table, as a range with a payback. "About fifteen desk hours one-off for the first five habits, which remove ninety to a hundred and twenty desk hours a month; the fixes pay back within the first month. The investments in hand are another forty hours across the quarter and pay back in two weeks to two and a half months." The answer is strongest when it asks for hours rather than money and for permission rather than budget. Redbrook's request was for forty desk hours this quarter and a decision from one approver about a laptop list. Nobody has to find money for that. Somebody has to say yes.

The third question is "what happens if we do nothing?", and the answer is the figure on the page, continued. "The hundred and sixty hours a month continue, which over the year is most of a technician's working time spent on things done twice, and the waiting line stays where it is." Say it plainly and without drama, because the figure is the drama and adding to it weakens it. A decision-maker who hears the cost of nothing stated calmly, with the number, will usually decide. Doing nothing has just been described as a choice with a price rather than as the default. The general craft of answering a manager's question in three sentences is in Explaining Technical Things to Management. For this meeting, the three sentences are the method, the hours, and the cost of nothing.

Presenting the Bad Numbers

Sometimes the economics show the desk wasting its own time, and the page says so, in the same type as everything else. A hundred and sixty removable hours a month says a quarter of the desk's time went on things done twice. A presenter who softens that line to protect the desk has made the whole page less believable, because the reader can see the softening. State it. Then state the fix and its payback on the next line. A desk that reports its own waste and its own plan to remove it is a desk that counts. That is a reputation worth more than the hundred and sixty hours. The general case of reporting a bad quarter honestly is in When the Numbers Are Bad. The economics version is no different: the bad line, the cause, the fix, the date.

Report the retired investments beside the live ones, for the same reason. Each is a half-sentence on page one. Each demonstrates that the payback figures for the surviving investments came from the same honest arithmetic. A page that shows only successes is read as marketing, by everyone, every time. A page that shows two stops and eight successes is read as accounting, and accounting is what the room came for.

Never present the bad numbers as somebody's fault, including your own predecessor's. "This is what we found when we counted" describes a situation; "this is what was allowed to happen" starts an argument about history that the decision does not need. The page is about the next quarter. The past is a baseline, and baselines do not have authors.

The Worked Example: One Page at a Fictional Desk

Here is the page Priya took to the second meeting, three weeks after the first, as prose; the template below is its shape. The question at the top was Ruth's own: "What does support cost Redbrook, and where could it cost less?" Line one: the desk has about six hundred working hours a month and spends about four hundred and seventy on tickets. About a hundred and sixty of those go on tickets fixed twice, requests handled without a ticket, and re-typing and meetings. This comes from the ticket records and a tally week. Line two: people who raised tickets waited somewhere between five hundred and a thousand hours last month. This is an estimate from forty tickets, in working hours, kept separate from line one. Line three: about 1.1 hours per ticket all in; a baseline, which will rise when self-service password resets remove the quick tickets, which is expected and good.

Line four: five habits cost the desk ninety to a hundred and twenty hours a month and requesters sixty to a hundred and twenty. Fixing them costs about fifteen desk hours and one decision from the laptop approver, and pays back within the first month. Line five: the six articles, the new-starter template and the disk-space sweep cost about forty desk hours this quarter. They should return twenty-five to sixty desk hours a month by next quarter, with payback between two weeks and two and a half months. Two earlier articles were retired as not paying back. The request: forty desk hours this quarter, agreement to the pre-approved laptop list, and a re-read of this page in three months. The method note: six lines, one per figure. No currency anywhere. Owen converted line one to money in the meeting, on his own pad, and was the one who read the result to Ruth.

The meeting took eighteen minutes. Ruth asked how Priya knew about the waiting line, got the two-sentence answer, and asked nothing further about it. The laptop list was agreed in the room because the approver was in it. The forty hours were granted with the phrase "that's not really a budget question", which is the phrase a page in hours is designed to produce. The nine-tab workbook was never opened again, and Priya has since described it as the most useful thing she ever built and never showed anyone.

A Short War Story: The Fourteen Tabs at Ashgrove Housing

At Ashgrove Housing, a desk lead called Callum spent a month building the case for a second technician as a fourteen-tab workbook. Every figure was converted to currency at a rate he had taken from an old planning document. The finance analyst, Rhiannon, arrived at the review with her own model built on the current rate, which was lower. She also had a different view of whether weekends counted in waiting time. Forty minutes of the meeting went on the rate and the weekends. The director closed it with "come back when you two agree". The decision slipped a quarter, then another, while the desk's measured backlog, which had justified the hire all along, grew by about a third. Callum rebuilt the case as one page in hours. The waiting line was a labelled range and the desk's own counted hours were the figure he defended. He gave Rhiannon the conversion to do in advance. The hire was approved at the next review in a meeting that lasted twenty minutes. The two quarters of delay had cost the organisation, by Callum's own cost map, roughly the hours the second technician would have provided in the first of them. A case presented in units the other side owns will be argued in those units, and the argument is the only thing the meeting will decide.

The Economics One-Pager

The template below is the page as a form, filled in from the five sheets in this series. These are the cost map, the cost-per-ticket read, the waiting-cost worksheet, the payback sheet and the habit cost table. Section A is the question in the reader's words. Section B is the five lines, each marked counted or estimated, each with its figure and the one-phrase source. Section C is the request or the observation, in hours and decisions, never money. Section D is the method note. Section E is the presenter's own preparation, filled in before the meeting and kept off the page. It contains the number to be defended and the rebuild, the concessions, and the three answers. Page two, the workings, is the five sheets themselves, brought and not presented.

SUPPORT ECONOMICS — ONE PAGE
Desk: ______________________   Period: ____________   Prepared by: ______________
For: ______________________   Conversion to currency, if wanted, by: ______________

A. THE QUESTION (in the reader's words)
   ________________________________________________________________________

B. THE FIVE LINES                                        counted / estimated
   B1 Desk hours a month ____; on tickets ____;
      on work done twice, unlogged requests and re-typing ____ ... ____________
      source: ______________________________
   B2 Requester hours waiting a month: ____ to ____ (working hours) ... ____________
      source: sample of ____ tickets, classified; never added to B1
   B3 Hours per ticket, all in: ____; reading: ______________________ ... ____________
      source: ______________________________
   B4 Habits priced: ____ habits costing the desk ____ to ____ hours a month
      and requesters ____ to ____; fixes cost ____ hours and ____ decision(s);
      payback ____ ............................................ ____________
      source: tally week and approval log
   B5 Investments in hand: ____ hours this quarter, returning ____ to ____
      hours a month by ____; payback ____ to ____; retired: ____ ... ____________
      source: payback sheets

C. THE REQUEST OR THE OBSERVATION (hours and decisions, never money)
   ________________________________________________________________________
   Re-read this page on: ____________

D. METHOD NOTE (one line per figure; page two holds the workings)
   B1 ______________________   B2 ______________________   B3 ______________________
   B4 ______________________   B5 ______________________

E. PRESENTER'S PREPARATION (not on the page)
   The number I will defend: ________________________________
   Its rebuild, aloud, in under a minute: ______________________________________
   Concessions to make first: 1. ______________  2. ______________  3. ______________
   "How do you know?" answer, two sentences: ___________________________________
   "What would it cost to fix?" answer, in hours: ______________________________
   "What if we do nothing?" answer, the figure continued: _______________________
   [ ] No currency anywhere on the page
   [ ] Every estimate is a range with a sample size; every count has a source
   [ ] B2 is on its own line and in no total
   [ ] The page is one page; the workings are in the folder

Priya keeps the filled-in sheet from each quarter in the same folder as the five it was built from. The second quarter's page, beside the first, was the first thing Ruth asked to see at the next review. The removable hours were down to about a hundred. The waiting line was down to between three hundred and six hundred and fifty. The two numbers sat on the same page in the same type, one counted and one estimated, with nobody arguing about either. Owen brought no model. He brought a pen.

Frequently Asked Questions