Presenting Monitoring Data in a Meeting Without Losing the Room
You have twelve minutes on the agenda of the management meeting, after the sales update and before the office move. You have a laptop full of graphs and a finding that matters. The link between the warehouse and the office is full every afternoon, and the scanner project will make it worse. Twelve minutes later the room has seen nine charts, asked what the vertical axis means, and moved on to the office move. Nobody decided anything.
A meeting is the worst place to present monitoring data and the best place to get a decision. And you cannot have the second without surviving the first. The people in the room cannot pause you, scroll back, or read the legend. They get one pass, at your speed, while thinking about the office move. Everything you show them has to survive those conditions.
This guide is about surviving them. You will learn to open with the answer. You will learn to give each point one chart with the point written on it. You will learn to choose charts a room can read in five seconds. You will learn to prepare for the three questions that always come and to take a challenge without losing your footing. And you will learn to present a good month without wasting anyone's time. There is a five-slide outline to copy and one rehearsal habit that fixes most of the rest.
In this guide
- The first sentence is the answer
- One chart per point, with the point on the chart
- A chart the room can read in five seconds
- The three questions, answered in advance
- Taking a challenge with grace
- The "we are fine" meeting
- The five-slide outline
- One rehearsal habit
- A war story
- The meeting checklist
- Where to go next
The first sentence is the answer
Your first sentence tells the room what you concluded and what you want. "The link between the warehouse and the office is full every afternoon, and I am asking for approval to get quotes for a faster one before the scanner project starts." Everything after that is support. If the projector fails, the fire alarm goes, or the chief executive leaves after ninety seconds, the decision is still in the room.
The instinct is to build up to the answer, because that is how you found it. You noticed the utilisation, checked the graph, ruled out a faulty counter, matched it against the complaints, and only then concluded. Presenting in that order asks the room to redo your investigation with less information and less time. It then asks them to arrive at your conclusion on their own. They will not arrive. They will arrive at the office move.
The rule that governs written reports governs spoken ones more strictly, because a listener cannot skip to the conclusion. The guide to reports people actually read makes the written case: lead with the answer, show the trend, stop. Out loud, the only difference is that you also have to stop talking.
Write the first sentence down before you make a single slide. If you cannot write it, you do not have a finding yet; you have a graph. (Graphs do not become findings by being shown to more people.) Go back to the data until the graph has grown a sentence, then build the slides to support the sentence.
One chart per point, with the point on the chart
Each point gets one chart, and the point is written on the chart as its title. Not "Interface utilisation, Gi1/0/48, 90 days" but "The warehouse link is full for three hours every afternoon, up from one hour in the spring." The room reads the title, glances at the line to see that it agrees, and is ready for your next sentence. A chart titled with its data source is a chart the room has to decode, and decoding is where you lose them.
Two charts for one point halves the attention each gets. Three points on one chart means nobody knows where to look. If a chart needs a legend to be understood, it has too many lines on it. A legend is a confession that the chart could not explain itself. Split it, or cut the extra lines and keep the one that carries the point.
Put the limit on the chart as a line and label it in words. The word "full" written beside a horizontal line at the top of a utilisation chart is worth more than the whole axis. Draw the line where you would call the link full. Mark the date where the trend crosses it. A room that can see the line, the limit, and the crossing date has understood the finding before you finish the sentence. Understanding before the sentence ends is what "not losing the room" means.
Keep the device names off. The room does not know what Gi1/0/48 is and does not need to. "The warehouse link" is the same thing, said in a way the operations director will repeat correctly afterwards. Speaking Business, Not Packets covers the vocabulary, and the chart follows the same rules as the sentence.
A chart the room can read in five seconds
A meeting chart has five seconds to land, because that is roughly how long the room looks at the screen before it looks back at you. The test is simple. Show the slide to a colleague who has not seen it, count to five, and ask what it says. If they read the title back to you, it works. If they ask what the axis is, it does not.
Most monitoring findings fit one of five shapes, and each shape has a chart that reads fast. Anything not in the table needs a reason to be in the room.
| What you want to show | Chart | Why it reads fast |
|---|---|---|
| Something is getting worse over time | One line, the limit drawn on and labelled, the crossing date marked | The eye follows the line to the limit |
| Something happened, and when | A timeline: one bar per outage on a date axis, height for duration | Dates are the unit the room already thinks in |
| A fix worked | Two bars, before and after, same scale | One comparison, nothing to read |
| How much of a limit is used | One bar against a labelled limit | It is a fuel gauge, and everyone can read a fuel gauge |
| Which of several things is worst | Ranked horizontal bars, worst at the top, no more than five | The top bar is the point |
Avoid the charts that need explaining. Pie charts, because nobody can compare two slices by eye. Stacked areas, because nobody can tell whether a layer grew or the one beneath it did. Two vertical axes, because half the room will read the wrong one. Anything with more than four lines, whatever the reason. The dashboard in your console can afford density, because you look at it every day. And the manager's dashboard is the screen version of this discipline, and it exists for the same reason.
The numbers on the chart should be the ones the room uses. Hours, days, people, and dates. Not percent of interface capacity, not milliseconds, not bits per second. If the monitor gives you the wrong unit, convert it before the chart and say in a footnote how you did. "Hours above 85 percent, from five-minute samples" is a footnote. It is not a title.
Tip: put the date range and the source on every chart, in small type at the bottom. Nobody in the room will read it, and the one person who asks "when was this?" will be answered by pointing. That is the entire reason it is there.
The three questions, answered in advance
Every presentation of monitoring data gets the same three questions, in some form, from someone. How sure are you? What does it cost? What happens if we wait? They are good questions and the ones you would ask. Because they are predictable you can have the answers on a slide before anyone asks. The slide is hidden after the last one, and you go to it only when the question comes. A hidden slide produced on demand looks like preparation. The same slide shown unasked looks like padding.
How sure are you?
"How sure are you?" is a question about evidence, and the honest answer names the kind you have. "It has happened three times; here are the dates" is the strongest. "This is ninety days of readings taken every five minutes, and the pattern repeats every working day" is strong. "The line is pointed at the limit, and I drew it through a full year rather than a quiet quarter" is honest about a projection. Framing Monitoring Findings as Risk and Cost calls these the three likelihoods. Say which one you have; the room trusts you more for keeping them apart.
What does it cost?
"What does it cost?" means both the fix and the problem, and the hidden slide carries both as ranges with the inputs shown. The fix is a quote from procurement or your own hours. The problem is the organisation's own arithmetic: people idle, times hours, times what an hour costs, with a labelled guess from sales for lost orders. Never quote an industry figure. The finance director in the room has heard them all and believes none of them.
What happens if we wait?
"What happens if we wait?" is the question the trend chart exists to answer, and the answer is a date and a consequence. "At the current rate the link is full all afternoon by the spring. With the scanner project it is full by the end of the quarter. When it is full, dispatch slows and the scanners time out." The date comes from one line of arithmetic, which belongs in the slide's footnote:
months until the limit = (limit - current level) / growth per month Spreadsheet: limit in B1, this month's level in B2, growth per month read from the graph in B3: =ROUNDDOWN((B1 - B2) / B3, 0) A result of 4 means "four months at the current rate".
Say "at the current rate" out loud every time, because the rate is the part you are least sure of. A project, a new team, or a quiet summer changes it. A date given without the caveat becomes a promise the room will remember better than you do.
The hidden slide holds all three answers in that order, two lines each. When the question comes you go to it, read the two lines, and come back. The room learns that you had thought about it, which is worth more than the answer itself.
Taking a challenge with grace
Someone in the room will disagree, and the presentation is judged on what you do next. The operations director says the warehouse has never complained. The finance director says the last upgrade did not fix anything. A challenge is not an attack; it is the room testing whether the finding survives contact with someone who was not there when you found it. Treat it as data. It usually is.
The method is to agree with whatever is true, add what they did not have, and return to the evidence. "You are right that nobody has logged a ticket. The scanner timeouts show up in the error log, not in tickets, because the pickers retry and move on. Here are the dates they cluster on." You have not conceded the finding and you have not called anyone wrong. The chart does the disagreeing, which is what it is for.
If you do not know, say so and say when you will. "I do not have that number here; I will send it by Thursday." A made-up answer in a meeting is found out within the week, and after that every number you show is checked. Not knowing one thing costs you nothing. Being caught inventing one costs you the year.
Do not argue with the person who signs. If the finance director is unconvinced after the evidence and the hidden slide, the meeting is not where you win. Ask what evidence would change their mind, write it down in front of them, and go and get it. That question ends more arguments than any chart, because it turns "no" into a specification.
The "we are fine" meeting
Some months nothing is wrong, and you should present anyway, briefly. Two minutes: "Nothing stopped anyone working this month. The one thing I am watching is the shared drive, which fills in the spring; I will bring options next month. I need nothing." Then sit down. The room has just learned that when you do stand up for twelve minutes, something is worth twelve minutes.
A good month is evidence too. It is the month the second internet line went live and nothing happened when the first one dropped for an hour. That sentence, said out loud with the date, is the only return-on-investment report the finance director will ever remember. Say it. The guide to reporting incident metrics to management covers the quarterly version of "things are getting better". The monthly version is one sentence with a date in it.
The other reason to present in a fine month is that absence is noticed. An IT slot that only appears when something is wrong becomes the slot people brace for. Braced is not the frame of mind you want them in when you finally do need money. (The reverse also holds: the slot that always overruns becomes the slot people schedule calls over.)
The five-slide outline
Five slides is enough for one finding and one decision, and it fits inside ten minutes with time for questions. The outline below is the template; the sixth slide is the hidden one.
SLIDE 1 THE ANSWER
Title: the finding in one sentence, in business words
Body: what you are asking for, and the date you need it by
SLIDE 2 THE EVIDENCE
Title: the point, as a sentence
Body: one chart; limit drawn on and labelled; crossing
date marked; source and date range in a footnote
SLIDE 3 WHAT IT MEANS
Title: who is affected and how
Body: three lines at most: people, work, money or risk
SLIDE 4 THE OPTIONS
Title: "Three ways forward"
Body: do nothing / patch / fix properly
one line each: cost, what it buys, how long it lasts
SLIDE 5 THE DECISION
Title: what you recommend
Body: the decision needed, from whom, by when;
what happens after that date if there is no decision
SLIDE 6 (hidden) THE THREE QUESTIONS
How sure: kind of evidence, sample count, dates
What it costs: fix (quote) and problem (range, inputs shown)
If we wait: the date it becomes acute, and the consequence
The options slide is where most presentations get ambitious and most rooms get lost. Three options, one line each, and the recommendation on the next slide rather than buried in this one. When the ask is large enough to need a written case, Building a Budget Case From Monitoring Data is the long form and this slide is its summary. Do not try to deliver the long form across a projector.
One rehearsal habit
Rehearse once, out loud, to someone who does not know the subject, with a timer running. Not in your head, where everything sounds fine. Out loud, where the sentence with three subordinate clauses falls over and the slide you thought was clear gets "what is that line?" The listener does not need to understand networks. They need to be able to tell you which slide they stopped following.
Time it, then cut a third. A rehearsal that runs ten minutes becomes fourteen in the room. That is because the room asks questions and the projector needs a moment to form an opinion of your laptop. If the slot is ten minutes, rehearse to seven and keep the cut material for the hidden slide.
I learned this from a presentation in which I reached slide four of nine when the chair said "let's pick this up next month." Next month had the office move in it. The finding waited a quarter, and so did the decision.
A war story
The network administrator at Harbourview Clinic was given ten minutes at the management meeting to explain why the clinic's internet line needed upgrading. The first slide was a ninety-day utilisation graph with fourteen lines on it, one per interface. The lines were in colours chosen by the monitoring tool, with a legend down the right-hand side in eight-point type. The practice manager asked which line was the internet. The administrator said the orange one, and two people asked "which orange?" Six minutes went on finding the line and three on the vertical axis. The chair moved to the car park resurfacing with a minute to spare. The upgrade was raised again the following quarter, by which time the line had been full every afternoon for three months. And the minutes recorded that the earlier presentation had been "detailed."
Imagine one line, titled "The internet line is full every afternoon from two o'clock," with "full" drawn across the top and the date marked. It would have taken five seconds and left nine minutes for the decision. Fourteen lines took ten minutes and left nothing. The number of lines on a meeting chart is not a measure of thoroughness. It is a measure of how much of the meeting you are about to spend explaining them.
The meeting checklist
Run the deck through this list the day before. Anything that fails goes on the hidden slide or in the bin.
- The first sentence states the finding and the ask, and it was written before any slide existed.
- Each point has exactly one chart, and the point is the chart's title.
- Every chart passes the five-second test with someone who has not seen it.
- Limits are drawn on the charts as labelled lines, with the crossing date marked.
- No chart has more than four lines, a legend, a second vertical axis, or a device name.
- Units are hours, days, people, and dates, converted before the chart, with the conversion in a footnote.
- The hidden slide answers how sure, what it costs, and what happens if we wait.
- Costs are ranges built from the organisation's own figures, with the inputs shown.
- The rehearsal was out loud, timed, and cut to two-thirds of the slot.
- A good month is presented in two minutes, naming the one trend being watched.
Where to go next
The sentence on slide one is built the way Speaking Business, Not Packets builds it. And the ranges on the hidden slide come from Framing Monitoring Findings as Risk and Cost. The written version the room can take away is The Monthly One-Pager. When the options slide grows into a request for money, Building a Budget Case From Monitoring Data is the long form. All of them live under Translating Monitoring Data for Management.
Put this into practice with SysPeep Monitor
The chart on slide two is one you already have, if the monitoring keeps history. SysPeep Monitor builds dashboards and reports from one web console over what it collects by SNMP, WMI, and SSH. So a ninety-day trend for one link is a report rather than a spreadsheet. And scheduled email reports mean the evidence arrives before the meeting does. Check the product documentation for the report types available. The free trial installs on any Windows machine in minutes.
Download Free TrialLearn About SysPeep