When Leadership Says No: Running Well on the Budget You Have and Asking Again
You built the case. The dividend was honest, the story landed, the budget line had all five parts and a trimmed alternative, and the answer was still no. Sometimes it is a quiet no, in an email, with "this year" in it. Sometimes it is a line item that simply is not in the approved sheet. Either way, the licence lapses in a few months and the renewal quote is now a piece of paper. You have to decide what kind of year this is going to be.
It is going to be a normal year, with less. That is the whole of this guide. And the reason it is worth three thousand words is that most administrators do one of two things after a no. Both are wrong. They sulk and let the monitoring decay to prove a point nobody is watching them prove. Or they re-ask immediately with the same slides, which I have done. It went about as well as the first time, only shorter.
By the end you will know how to accept the no in a way that sets up the next ask. You will know how to keep monitoring running well on what you have. You will know which three things you protect when everything else is cut, and how to keep the evidence file growing. You will know what the re-ask triggers are and how to use one without saying "I told you so". And you will have a one-year plan you can fill in the week the answer comes back.
In this guide
- A no is a decision about this year's money
- The graceful acceptance
- Running well on what you have
- Protecting the essentials when cutting
- Keeping the evidence file growing
- The re-ask trigger
- The one-year plan template
- A war story
- The letter you do not send
- The after-a-no checklist
- Where to go next
A no is a decision about this year's money
A no is a decision about this year's money, not a verdict on you, your case, or the monitoring. Budgets are finite, other lines wanted the same money, and some of those lines had a person behind them with an older relationship and a louder story. The finance director did not conclude that monitoring is worthless. They concluded that this year, with these numbers, something else came first. That is a decision with a date on it, and decisions with dates get revisited.
It helps to know what kind of no you got, because they are not all the same. "Not this year" is a scheduling no, and the re-ask is a calendar entry. "Not at that price" is a negotiating no, and the trimmed line from the budget line guide is the reply. "I do not see the value" is an evidence no, and the year ahead is about building the evidence they did not see. Ask which one you were given. Most leaders will tell you, and the answer decides everything below.
The graceful acceptance
Accept the no in three moves, in the same meeting if you can and by the end of the week if you cannot. Thank them, ask what would change the answer, and write the answer down. All three are for the next ask, not this one.
Thank them, because they read the case, and reading a case is not nothing. "Thanks for going through it properly" costs you nothing and buys you a leader who reads the next one. Then ask the only question that matters: "What would change the answer?" You will usually get one of a few replies. It may be a lower number, a different year, or an incident that proves the point. It may be a comparison with a peer, or a new leader's priorities. Whatever it is, that is the re-ask trigger, and they have just told you what it is.
Write the answer down, that day, in the evidence file, in their words. Include the date, who said it, what was asked for, what was declined, and what they said would change it. This is the single most important act of the year. A re-ask that begins "in March you said an incident would change the answer; here is the incident" is a different conversation from one that begins "I still think we need this". (Write it in the file. Not on the whiteboard behind the finance director.)
RECORD OF A NO Date: [date] Asked for: [the line, total, and the trimmed alternative] Answer: [no / not this year / not at that price / value unclear] Given by: [name, role] Their reason: [their words, as close as you can get] What would change it: [their words] Re-ask trigger: [the concrete event that matches the above] Next review: [date, usually the next budget round or the renewal]
Running well on what you have
The year after a no is a year of running monitoring well with less, and "well" is the operative word. Monitoring that is allowed to decay proves nothing except that it was not looked after. A leader who walks past a dashboard showing last month's data will draw exactly that conclusion. The decision is theirs; the standard is yours.
What you have is more than it looks. Check the product documentation for what happens to collection and alerting when a licence lapses, because it varies. Plan around the answer rather than discovering it. The built-in tools of the operating system, the switches' own counters, and the small free utilities every administrator carries are a real second layer; the essential utility belt lists them, and the monitoring on a budget pillar is the year's reading. Where spending actually helps is worth reading twice. It tells you which of the things you lost were the ones that mattered. That is the list for the re-ask.
A lab costs almost nothing and keeps your hand in. An old desktop and a free trial of a monitoring product make a place to keep skills current. You can test a threshold idea and produce the occasional graph for a story. And nobody has ever been refused permission to install a trial on a machine that was going to the skip. Keep it honest: a trial will expire, and it is not the production monitoring. It is where you practise for the year you get the budget back.
Protecting the essentials when cutting
When the line is cut and you are choosing what to stop, three things are protected before anything else is considered. They are the monitoring server's own disk and backups, the alerting path, and the evidence file. Everything else can be reduced. These three cannot, because each one is the thing that makes the re-ask possible.
The server's disk and backups, because the history is the asset. A year of graphs and alert logs is the evidence for every story and every dividend row. And it lives on one disk on one server nobody is now paying much attention to. Keep the backups running and the disk from filling; the guide on backing up the monitoring system is the afternoon's work that makes the rest of the year safe. Run this once a month and write the two numbers in the plan:
# Monthly essentials check: free space on the monitoring data drive,
# and the age in days of the newest backup file. Adjust drive and path.
Get-PSDrive -Name D |
Select-Object Name, @{n='FreeGB'; e={[math]::Round($_.Free / 1GB, 1)}}
$newest = Get-ChildItem "E:\Backups\Monitoring" -File |
Sort-Object LastWriteTime -Descending | Select-Object -First 1
[math]::Round(((Get-Date) - $newest.LastWriteTime).TotalDays, 1)
The alerting path, because a fault found by a user is the one outcome the year cannot afford. If the licence lapse reduces what can be alerted on, keep the alerts for the few things whose failure stops the business. In that case, let the rest go quiet deliberately, with a note in the plan saying which ones and why. A heartbeat check that tells you the monitoring itself has stopped is the one alert you add this year rather than remove.
The evidence file, because it is the next ask. Its own section follows. The short version is that the file does not lapse when the licence does. Never let it. Not for a quarter. Not because "nothing happened".
Tip: write the list of what you stopped monitoring, with the date, in the plan and in the evidence file. When one of those things fails, the entry is the re-ask, already dated, in your handwriting, from before the failure. That is the difference between evidence and hindsight.
Keeping the evidence file growing
The evidence file is the folder where you keep the dated record of what the network did. It holds availability figures, incident timelines, capacity catches, provider disputes, and now the record of the no. The uptime evidence file guide describes how to build and keep one. This guide's addition is that in the year after a no the file has a second purpose. It makes the next ask a different conversation from the last one.
Record four things this year, every month, on a fixed day. Record the incidents, with how each was detected and whether a monitoring item you cut would have caught it. Decide that at the review while it is fresh, not in budget week. The out-of-hours alerts and calls, counted, because the human cost of less monitoring is part of the case. The capacity catches you still made, and the ones you could not because the trend history stopped. And the running total of outage minutes, in the same cost classes the dividend used, so next year's dividend is built on the same basis as this year's.
A query keeps the first of those honest, if your incident records carry the fields. The would_have_alerted column is a judgement. It is only credible if it was recorded at the post-incident review, so add the column and fill it at the time:
-- Incidents since the no that users found first, where a monitoring -- item we stopped would have caught it (judged at the review, not later). SELECT incident_id, opened_at, service, minutes_down, cost_class FROM incidents WHERE opened_at >= '[date of the no]' AND detected_by = 'user' AND would_have_alerted = 'Y' ORDER BY opened_at;
Each row is a sentence in the re-ask, and each has a ticket behind it. Three rows and a cost class are a case. Zero rows is also a finding, and an honest file records it. A year in which nothing the cut would have caught went wrong is a year in which the leader was, on the evidence, right. Saying so is what makes them believe you the year they are not.
The re-ask trigger
A re-ask trigger is an event that changes the answer to the question you were asked. You re-ask when one happens, not before and not on a schedule of your own frustration. There are four, and the leader usually named one of them when you asked what would change the answer.
- An incident the cut would have caught. The strongest trigger, and the one to use most carefully. The evidence is the query above, the ticket, and the dated note from before the failure listing the item you stopped.
- A growth step. A new site, a large batch of devices, a service the business now depends on. The network the no was decided for no longer exists, so the decision is due again.
- A renewal or a refresh. The licence anniversary, the monitoring server's replacement year, or a change to the product's terms. Money is being discussed anyway, and a line that arrives with a year of evidence is a different line.
- A new leader. A new finance director or IT director has not said no. They have said nothing. The case, the dividend, and the file are a first conversation, not a repeat.
Re-asking without "I told you so" is a matter of tense and subject. The subject is the incident, never the decision. And the tense is present: "Here is what happened in June, here is what it cost, here is the item that would have caught it, which is on the list we agreed to stop in March." Every fact in that sentence is in the file, dated, and the leader can draw the conclusion themselves. They will do that more willingly than if you draw it for them. Then present the same line as before, updated, with the trimmed alternative, and "what would change the answer?" already answered in their own words from the record. The guide to building a budget case covers the structure. In a re-ask the only change is that the evidence section is a year longer and partly in the leader's handwriting.
The one-year plan template
Fill this in the week the no arrives, one row per quarter. It is a plan for running well with less, and every cell is something you will actually do. The last column is the event that would send you back to the leader in that quarter; leave it as "none expected" when that is the truth.
| Quarter | What we keep | What we drop | What we record | Re-ask trigger |
|---|---|---|---|---|
| Q1 | Core alerting on [list]; server backups; monthly essentials check; evidence file | [Sensors, sites, or retention reduced, with dates] | Record of the no; list of items stopped; incident log with detected-by and would-have-alerted | Growth step: [known project], or none expected |
| Q2 | As Q1; lab install on [machine] for skills and graphs | [Further trims if the lapse requires them] | Out-of-hours alerts and calls; capacity catches made and missed; outage minutes by cost class | Incident the cut would have caught; otherwise none expected |
| Q3 | As Q2; review the kept-alert list once | [Nothing new unless forced; note if forced] | Half-year dividend on the same basis as last year; one before-and-after story if one exists | Renewal quote arrives: prepare the line and the trimmed alternative |
| Q4 | As Q3 | [Nothing new] | Full-year evidence pack: incidents, minutes, dividend, the record of the no with its trigger matched or not | Budget round: re-ask with the file, or record that the no was right this year |
The plan is also what you show the leader when they ask, in the second month, how things are going. "Running to the plan, here it is" builds a reputation for the year you need one. A plan with "none expected" in the last column is more credible than one where every quarter predicts disaster, because it looks like a plan and not a grievance.
A war story
Greyfriars College had its monitoring renewal declined with the words "show me one thing it would have stopped". The administrator wrote that in the file with the date and the director's name. Over the following year the file collected three incidents that users found first. Each had a note from the review saying which stopped item would have caught it and how many minutes it ran. There was also one quarter with nothing at all, which was recorded too. At the next budget round the administrator opened with the director's own sentence. They read the three rows and the quiet quarter, and put the same line on the table with the same trimmed alternative. The director asked whether the quiet quarter meant the case was weaker. The administrator said yes, slightly, and that it was in the file for that reason. The line was approved. The director's only comment was that it was the first time anyone had brought her something she had said back to her in writing and used it fairly.
The letter you do not send
There is a letter every administrator writes in the week after the incident the cut would have caught. It begins "As I said in March". It lists every fault since the decision, each followed by the phrase "which would have been caught", in a font that gets slightly larger as the list goes on. It closes by asking, in careful language, for an acknowledgement, which is a formal word for an apology. It is a good letter, in the sense that it is accurate. It should be written, read once, and deleted. Every sentence in it is about the decision and the person who made it, and none is about the network.
The version you send is the Q3 row of the plan. It covers the incident, its cost, the stopped item, and the date on the list, in the present tense, with the file behind it. It says everything the letter said and leaves the conclusion to the reader. That is the only way a conclusion ever gets adopted by a person with a budget.
The after-a-no checklist
Work through this in the week the answer arrives, and again each quarter. Every item is a yes or no.
- You know which kind of no you were given: scheduling, negotiating, or evidence.
- You thanked them and asked what would change the answer.
- The record of the no is in the evidence file, dated, in their words.
- You have checked what the product does when the licence lapses and planned around it.
- The list of what you stopped monitoring is written down with dates.
- The monitoring server's backups run and the monthly essentials check is on the calendar.
- The kept-alert list covers what stops the business, and a heartbeat check covers the monitoring itself.
- Incidents are logged with detected-by and would-have-alerted, filled in at the review.
- Out-of-hours alerts, capacity catches, and outage minutes are recorded monthly in last year's cost classes.
- The one-year plan is filled in, with an honest "none expected" where that is true.
- The re-ask trigger is named, and the re-ask will be about the incident, in the present tense.
- The letter has been written, read, and deleted.
Where to go next
A no is a date, not a verdict. Thank them, ask what would change it, and write the answer down in their words. Then run the year well with less. Protect the server's disk and backups, the alerting path, and the evidence file. Let everything else be reduced deliberately, with dates. Record what happens, including the quarters where nothing does. Re-ask when a trigger fires, in the present tense, about the network, with the file open. The second ask is a different conversation because you made it one.
The re-ask is built from the same pieces as the first. Those pieces include what monitoring spend actually buys for the plain sentences. They include the downtime math guide for the dividend. They also include before-and-after stories for the one story that lands. The monitoring maturity model is useful reading in a lean year, because it shows what improves without money. All of them are part of Justifying Monitoring Spend to Leadership.
Put this into practice with SysPeep Monitor
A lab is the cheapest way to keep your hand in during a lean year. The free trial of SysPeep Monitor at /download/ installs on one Windows machine as a central service. Dashboards, reports, and alert history share a single web console. That is enough to practise thresholds and produce the occasional graph for a story. It is a trial and it will expire; it is not the production monitoring, and this guide would not pretend otherwise.
Download Free TrialLearn About SysPeep