Office analytics

The Office page covers one office at a time and answers what is actually happening in it: how much of the space is used, what would change if the office were sized differently, how many bookings go unused, and how many people come in. Select FlowAnalytics in the left menu, then Office.

The page has four tabs, each answering a different question:

The filter bar, saved filters, targets, display options, AI insights and the export functions work the same way here as everywhere else, and are described in Filters, targets and exports. What each figure measures is defined in Metrics reference. This article covers what each tab shows and how to read it.

Office Utilization

This is the tab to open when the question is whether the office is the right size. It reports utilization three ways, sets each against a target, and then breaks the figure down into what was booked and what was actually used.

Set what the tab covers

The filter bar offers Office, floor, zone, Asset type, Time period, Days, Time and Advanced.

Set the asset type before reading anything. A figure covering desks and meeting rooms together is an average of two things that behave differently, and it will tell you nothing useful about either. Look at desks, then look at rooms.

The key figures

Three gauges report utilization, each with its target and the change against the preceding period:

• Average utilization. The average share of space in use across the period. Use it to judge whether the office is the right size.
• Average daily peak utilization. The average of each day’s busiest moment. Use it to judge whether there are enough desks for a normal day.
• Peak utilization. The highest point reached in the whole period. Treat it as the worst case.

Reading the three together is what makes them useful. Average utilization of 43% with an average daily peak of 47% describes an office that is evenly and lightly used. The same 43% average with a peak of 90% describes an office that is under pressure at the busiest hour and empty for much of the rest of the day, and those two offices need different decisions.

Alongside the gauges sit three cards translating utilization into money and floor area: Potential savings / year, Potential space savings / year and Potential CO₂ reduction / year. These depend on the area and space cost recorded for the office. Until those are entered, each card shows Setup required with a link to the office settings. How the three are calculated is described in Metrics reference.

Asset utilization chart

The chart under the gauges is where the real finding usually is, because it separates bookings from presence. Each bar is split into four series:

• Booked and used. The reservation was honored.
• Used without booking. Somebody used the space without reserving it.
• Booked but unused. The reservation was never used. This is the no-show volume, and it is capacity that nobody else could book.
• Booked (no sensor data). The asset has no sensor, so presence could not be measured either way.

A line across the bars shows peak utilization for each period. Select Monthly, Daily or Hourly to change the granularity, select an item in the legend to hide or show a series, and where the chart says so, select a bar to filter the rest of the page by it.

Check the Booked (no sensor data) share first, before drawing any conclusion from the page. Where that share is large, utilization is largely a record of what people reserved rather than what they did, and a high figure may simply mean the office books well rather than that it is full.

Used without booking is worth attention for the opposite reason. A high share suggests booking is not needed in that area, or that people have stopped trusting it and now just sit down. Either way the booking rules are not matching behavior.

Hourly utilization

Average utilization by hour, which gives the shape of the working day. This is where you see whether the office fills mid-morning and empties after lunch, or holds steady all day.

The hours shown follow the Time filter, so hours excluded from the selection are absent from the chart. Set the filter to Working Hours for a picture of the whole day, or to Peak Hours when the question is whether there is enough space at the busiest time.

Weekday utilization

Average utilization by day of the week. In a hybrid office this is often the most actionable chart on the page, because the gap between the busiest and quietest day is what a policy can move. An office at 65% on Tuesday and 32% on Friday does not need more space; it needs the demand spread.

Zone utilization

Average utilization by zone, so a quiet corner of an otherwise busy floor becomes visible. Select the sort control to order the zones by utilization.

Zones only appear here if they have been defined for the office under Company. An office with no zones set up shows a single bar.

Floor utilization

Average utilization by floor, with the percentage shown alongside each bar. Select the sort control to reorder.

This is the chart behind a decision to release a floor. A floor well below the rest of the building is the obvious candidate, but check the weekday and hourly charts for it first, because a floor that is quiet four days a week and full on Wednesday still has to be somewhere.

Least utilized assets

Individual assets ranked by utilization rate, with the rate shown for each. The list opens with the least used first and is paginated, so an office with more than ten assets has further pages. Select the sort control to reverse the ranking and put the most used assets at the top.

Both directions answer a question. Least used first finds assets to remove, fix or remap. Most used first finds the desks and rooms people compete for, which is often a better guide to what to add than an average is.

The bottom of this list is worth checking asset by asset rather than reading as a statistic. An asset at or near 0% is often not an unpopular desk but a broken or unmapped sensor, a desk that no longer physically exists, or one that has been left in the inventory after a refit. Each of those inflates the denominator of every utilization figure on the page, and removing them is usually the quickest improvement available to the data.

Office Utilization Simulator

The simulator answers a different kind of question from the rest of FlowAnalytics. Every other dashboard reports what happened; this one asks what the same office would look like under different conditions, so it is the tab to use before committing to a change rather than after it.

The three scenario controls

Three cards each show the office’s current value and a scenario value you set with a slider:

• Assets. How many assets do you want to provide? Use it to test releasing desks, or adding them.
• Headcount. How many employees will use this office? Use it to test growth, or a team moving in or out.
• Office policy. How many days per week in the office? Use it to test a change from two days to three.

Select Reset on a card to return it to the current value.

The figures below the sliders then recalculate for the scenario, using the same gauges and charts as the Office Utilization tab, so you can read the effect in the same terms you already use.

How to use it

1. Set the filters so the tab covers the office and asset type you are considering.
2. Note where the utilization figures sit before you change anything.
3. Move one slider, and read the change in average utilization, average daily peak and peak.
4. Return that slider to its current value before testing the next one.

Change one variable at a time. Three sliders moved together produce a number nobody can explain, and the point of the exercise is usually to find the single change that gets utilization into its target range.

The most common use is the one the numbers make uncomfortable. An office at 43% average utilization looks like it could lose a third of its desks, until the daily peak is checked in the scenario and turns out to land above 90%, which means people would be hunting for a desk on ordinary Tuesdays. The average tells you what to consider and the peak tells you what it would feel like.

Where the current values come from

Each card shows a Current value, and the three come from different places:

• Assets. The number of assets in the current filter selection, the same figure as the assets selected count above the cards.
• Headcount. The Total headcount recorded for the office under Company → Offices. It is also shown as Headcount in the Company Information block at the foot of the dashboard settings dialog.
• Office policy. The days per week set for the office under Company → Offices.

The assets figure following the filter selection has a practical consequence: narrow the filter to one floor and the simulator is simulating that floor, not the office. That is usually what you want when a single floor is under review, but check the assets selected count before reading the result.

An office with no headcount recorded cannot be simulated meaningfully, since two of the three controls are then working against a blank.

A scenario is remembered

Scenario values stay set when you leave the tab and come back, which is useful while you are working through a case over several sittings.

It also means the tab may not be showing actuals when you next open it. Before reading the figures as fact, or showing the tab to somebody else, check each card for a Scenario value that differs from Current, and select Reset on any that do. A screenshot of a scenario is indistinguishable from a screenshot of reality once it is in a slide deck.

No Show

A no-show is a booking that was never used, and it is the most directly actionable thing on the Office page. Unused bookings block capacity that other people wanted, so a high no-show rate means the office is short of space it already has.

Set what the tab covers

The filter bar offers the same filters as Office Utilization, including Asset type. Desks and rooms behave differently here too, so look at them separately rather than leaving asset type unselected. This is a good tab to keep saved filters for, since a desks view and a rooms view over the same period are what you will return to.

The key figures

• No-show. The share of bookings that resulted in a no-show, counting both missed check-ins and bookings where no presence was detected.
• Late cancellation (0-1h). Bookings cancelled less than an hour before the start.
• Late cancellation (1-4h). Bookings cancelled one to four hours before the start.
• Recovery rate. The share of released assets that somebody else then booked or used.

Recovery rate requires check-in to be activated, since without check-in nothing is released. An office that does not use check-in has no recovery rate to report however high its no-show rate is.

Read the cancellation figures as the better version of the same behavior. A cancellation at least releases the space, while a no-show holds it to the end of the booking. Moving people from no-shows to early cancellations is usually a more achievable goal than eliminating either.

No-show over time

The trend chart shows the no-show share across the period, split into the same four categories as the cards. Select Monthly or Daily to change the granularity, and select an item in the legend to isolate a series.

One tall bar in an otherwise flat chart is usually an event rather than a trend: a company day, a public holiday somewhere in the selection, or a team that booked a floor and then met elsewhere. Check the date before treating it as a finding.

Missed bookings by type

This chart takes the four categories and splits each one by whether it came from a single booking or from a recurring event series.

The distinction matters because the fix differs. Single bookings are individual behavior, addressed by shortening advance booking windows or by automatic release. A large share from event series is a standing meeting that no longer happens, or a weekly desk booking somebody set up and forgot, and one conversation can clear a whole series at once.

No-shows by detection source

A donut showing which signal detected each no-show. Three segments can appear:

• Check-in & Sensor. Both signals were available on the asset, which is the most reliable configuration.
• Check-in. Check-in alone detected it, so the asset has no sensor.
• Sensor. The sensor alone detected it, which is what an office without check-in reports.

Read this as a data quality reading as much as a behavioral one, because each configuration misses something different. Check-in alone counts a booking as attended if somebody checked in and then left. Sensor alone catches genuine absence but has no record of intent, so it cannot distinguish a booking somebody abandoned from one they never meant to use. Assets with both are the ones whose figures you can rely on.

A no-show rate is therefore only as trustworthy as the mix in this chart. Where most of the volume sits under one of the single-signal segments, treat the headline rate as a floor rather than a measurement.

Missed bookings by department

No-show share by the department that made the booking, stacked by the same four categories.

This is the chart that turns a number into a conversation, because it identifies who to talk to. A department at 100% no-show share is not a discipline problem so much as a booking habit that nobody has questioned, often a recurring team booking that outlived the team.

The chart needs department information on the bookings, which depends on the identification level set for the system. Where that is set to store neither user nor department, this chart has nothing to show. See About FlowAnalytics for how that level works. Bookings with no department appear under N/A.

Recovery by release type

The share of released assets that get rebooked within the day, shown separately for each release type: late cancellations at one to four hours, late cancellations under an hour, and no-shows detected by a missed check-in.

This is the practical case for asking people to cancel rather than simply not turning up. An asset released four hours ahead stands a much better chance of being taken by somebody else than one released at the last minute, and the gap between the bars is that argument in figures.

Note that this chart counts rebooking within the day, while the recovery rate card counts released assets that were booked or used by someone else. The two answer slightly different questions, so they will not match exactly.

Attendance

Attendance counts people rather than assets, which makes it the tab to use for questions about an attendance policy. Utilization tells you whether the space is used; attendance tells you how many people came in, and those are different numbers that are easy to confuse.

Set what the tab covers

The filter bar here is shorter than on the other tabs. It offers Office, floor, zone, Time period, Days and Advanced, and has neither an asset type filter nor a time-of-day filter, because the tab counts people rather than desks and counts them by the day rather than by the hour.

The tab carries AI insights above the cards, which is often the fastest way in here: it names the busiest and quietest weekdays with their figures, and points out where the trend has moved.

The key figures

• Average in office. The average number of people per day in the office.
• Average remote. The average number working remotely.
• Average off work. The average number not working.
• Offices in selection. How many offices the figures cover, and how many countries.

Average in office is measured, from bookings, presence and check-in. Average remote and Average off work are not: employees set their own status in the Workplace app, so those two figures are declared rather than detected.

That difference decides how much weight each figure carries. Where people do not use the status feature, both cards read zero, and zero here means nobody said so rather than nobody worked remotely. Read remote and off work as a floor, and treat them as a useful signal only in an office where setting a status is an established habit.

It also means the three figures are not a split of the whole workforce and will not add up to headcount. Read them alongside Average in office rather than as shares of one total.

All three count whole days. There is no half-day option, so somebody who comes in for the morning counts the same as somebody there from nine to five. When the question is how much of the day the office was actually occupied, that is a utilization question rather than an attendance one.

Attendance over time

Daily attendance across the period, as three series: Office, Remote and Off Work. Select an item in the legend to hide or show a series, and use the slider under the bars to zoom into part of the period.

This is the chart to hold against a policy change. A policy introduced mid-period shows up here as a step rather than a slope, and if it does not show up at all, it has not landed.

Average attendance by office and by weekday

Two charts sit below the trend, both stacked by the same three series as the trend chart.

Average attendance by office gives average people per office across the period. It earns its place when several offices are in the selection; with a single office selected it is one bar, and the trend chart tells you more.

Average attendance by weekday gives the average attendance pattern across Monday to Friday.

The weekday chart is usually the one that changes a decision. Where Tuesday and Wednesday carry most of the week and Friday is nearly empty, the office is not too small; the demand is concentrated. Spreading it costs nothing in floor area, and it is the cheapest capacity you will find.

Reading the four tabs together

The tabs answer one question each, but a decision usually needs several of them. A worked sequence for the most common case, an office that looks too big:

1. On Office Utilization, filter to desks and check average utilization against average daily peak. A low average with a low peak is genuine slack; a low average with a high peak is concentrated demand.
2. Check the Booked (no sensor data) share in the asset utilization chart. Where it is large, treat the utilization figures as booking figures and be careful what you conclude.
3. On No Show, see how much of the booked time was never used. Unused bookings are capacity you already have.
4. On Attendance, check the weekday pattern. Concentrated demand is a policy question, not a floor area question.
5. On Office Utilization Simulator, test the change you are considering and check where the daily peak lands, not just the average.

Where the four disagree, the explanation is almost always the selection, the period, the asset type or the days and hours. Metrics reference has a chapter on that.