User analytics

The User page reports attendance and no-shows per named individual. It is the only part of FlowAnalytics that identifies people, and it is restricted separately for that reason.

Select FlowAnalytics in the left menu, then User. The page has two tabs:

Filters, targets, exports and the chart controls work as described in Filters, targets and exports, and what each figure measures is defined in Metrics reference.

Access and prerequisites

Two conditions have to be met before this page shows anything.

• The FlowAnalytics User Behavior role. The User page opens only for that role. Global Admin does not include it, so an administrator who sees every other dashboard will not see this one until the role is added.
• An identification level of user ID. The level set for the system decides whether the individual is stored with the data. At department level or lower these dashboards have nothing to report, even for someone holding the role.

Both are described in About FlowAnalytics. The identification level is set under Company settings by Flowscape and changed by contacting support@flowscapesolutions.com, so it is not something an administrator can switch on to answer a one-off question.

Before you use these dashboards

These two dashboards are different in kind from the rest of FlowAnalytics. Everywhere else the subject is a building; here it is a person, and that changes what the data can responsibly be used for.

Three things are worth settling before the figures are used in any conversation:

• What was agreed. Individual-level measurement is commonly covered by an agreement with employee representatives or a works council, and often by a stated purpose in a privacy notice. Check what your organization agreed to before using these dashboards for something else.
• What the figures actually measure. Attendance is presence in an office, from bookings, sensors and check-in. It is not hours worked, not output, and not availability. Somebody who spends the week on a customer site has low attendance and a full week of work.
• What the data cannot see. Attendance counts whole days, so a person in for the morning counts the same as one there all day. Presence needs a sensor, a booking or a check-in, so somebody working in an area with no coverage may not register at all.

The practical version of all three: use these dashboards to explain a pattern the aggregated dashboards have already shown, rather than to look for individuals.

Set what the dashboards cover

The filter bar on both tabs offers Office, floor, zone, Time period, Days, Department, Employees and Advanced. There is no asset type filter and no time-of-day filter, because both tabs count people by the day.

The Employees filter narrows the page to specific people. Where a review concerns one team, filtering by Department first and leaving employees unset is usually the better order: it keeps the comparison inside a group that shares the same working pattern.

The Employees filter lists every user with records in the last 180 days, which is a wider window than the period you are viewing. Somebody can therefore appear in the filter and have no data in the selected period.

User Attendance

This tab reports how often each person came to the office across the selected period.

AI insights

The tab opens with AI insights, which here reads the distribution rather than the total: how many employees had recorded attendance, how they group into bands from very low to very high, and who is highest.

That distribution is the useful part. An average of half a day per week could mean everybody comes in occasionally, or almost nobody comes in and one person is in most days, and the two call for entirely different responses.

As everywhere in FlowAnalytics, the insights are pre-calculated for the last 90 days and do not follow the period filter. On this page that is easy to trip over, because the insight text quotes a number of employees and an average that will not match the card above it whenever the period is set to anything other than the last 90 days. Compare like with like before treating a difference as an error.

The key figure

Average days in office gives the average days per week across the employees in the selection. It is the same metric as on Department Overview, at a different level of aggregation, so the two should agree when the selections match.

How often each person comes to the office

Average days per week in the office per employee, for the selected period, with a sort control to order the list and a download icon.

Read the shape before reading the names. A long tail of people at or near zero usually says something about the office or the policy; a handful of outliers in an otherwise even distribution is about individuals. Only the second is a case for a conversation with anyone.

Only people with recorded attendance above zero appear in the chart. Somebody who did not come in at all during the period is absent from it rather than shown at zero, so the chart cannot be read as a list of everybody. The insight text states how many employees had recorded attendance, which is the number to compare against your headcount.

That absence has several possible causes and the chart cannot tell them apart: long-term leave, a secondment, a role that is genuinely off site, or an office area with no sensor coverage. Check before assuming.

Who was in the office each day

A grid with one row per person and one column per day, where each highlighted square is a day that person was in the office. Days are grouped by month, a Total column gives each person’s day count for the period, and the grid can be downloaded.

This is the most detailed view in FlowAnalytics and the one to use when a pattern matters more than a total. Two people at fifteen days each look identical on the bar chart above; here you can see that one came in three days a week throughout while the other came in daily for three weeks and not since.

It is also the view to be most careful with, because it shows individual days rather than an average. Use it to answer a question that has already been framed, not to browse.

User No Show

This tab reports which individuals book space and then do not use it. Of the two tabs it is the more directly actionable, because a repeated no-show blocks a desk or room that somebody else wanted.

AI insights

The insight text names the office no-show rate and then the individuals whose booking behavior accounts for most of it, with their booking volume alongside their rate. Booking volume is the part to read carefully: a 100% no-show rate across twenty-four bookings and the same rate across two are different problems.

The key figures

• No-show. The share of this selection’s bookings that resulted in a no-show.
• Late cancellations. The share cancelled shortly before the start.

The card combines both cancellation windows into one figure, while the charts below split them into 0-1 hour and 1-4 hour. There is no recovery rate card on this tab, since recovery is a property of the released asset rather than of the person who released it.

Four per-employee charts: counts and rates

The tab reports the same behavior twice, once as totals and once as rates. All four charts are paginated, sortable and downloadable.

• No-shows per employee. Total no-shows in the period, stacked into No-show (No presence) and No-show (Missed check-in).
• Late cancellations per employee. Total late cancellations, stacked into Late cancellation (1-4h) and Late cancellation (0-1h).
• No-show rate per employee. No-shows as a percentage of that person’s total bookings, with the same two series.
• Late cancellation rate per employee. Late cancellations as a percentage of their total bookings.

Read the pair together, because each on its own is misleading. The counts favor whoever books most: somebody booking a desk daily and missing one in five appears above somebody who books twice a month and never turns up. The rates correct for that but flatter low-volume bookers, since two bookings and two no-shows is 100% and tells you almost nothing.

So the people worth a conversation are those high on both charts, and the insight text helps here by naming an individual with their rate and their booking volume side by side. Two hundred bookings at 80% is a different problem from four bookings at 100%.

The split between No presence and Missed check-in is worth a glance too. A person whose no-shows are all missed check-ins may be attending and not checking in, which is a habit worth correcting before it is treated as absence.

Employees with no no-shows or late cancellations do not appear in these charts, and bookings with no employee recorded are grouped under N/A.

Working from the aggregate down

The order in which these dashboards are opened matters more than anything else in this article. A sequence that keeps the answer proportionate:

1. On the No Show tab of the Office page, establish that there is a problem and how large it is.
2. Check No-shows by detection source there. Where much of the volume is detected by one signal only, the rate is unreliable and individual figures inherit that unreliability.
3. On Missed bookings by department, see whether it is concentrated somewhere.
4. Only then open User No Show, to find the specific bookings behind it.

The same applies to attendance: Department Overview shows whether a policy is landing, and User Attendance explains a department that stands out. Starting at the individual level tends to produce a list of names without a question attached to it.

Exporting from these dashboards

Select Export at the top right to produce the dashboard as a PDF, as described in Filters, targets and exports.

An export from this page is a document listing named individuals and their attendance or booking behavior. Treat it as personal data: it should go only to people who already have access to the same figures in the portal, and it should not outlive the question it was produced for.