Time and Activity: Trends and variance
Objective. Read an attendance or utilisation trend over time and separate a real change in behaviour from a change in scope, calendar or coverage.
This screen answers 'is this getting better or worse over time, and by how much?' It plots attendance, available hours or utilisation across periods so a manager can see direction and spot a break in the pattern.
See it in the demo
What this screen answers
A Sales or CSM person opens Trends when the question has moved from 'what happened yesterday' to 'what is the direction'. It shows how a measure moves across days, weeks or months for a chosen team or scope, so a manager can tell steady state from drift and see where a step change begins. In a demo it supports the 'you will see patterns you cannot see in a spreadsheet' story.
The task
Choose the measure first (attendance, available hours or utilisation - they trend differently), then set the scope to one team or account and a period long enough to show normal variation. Read the baseline band before the latest point: what does an ordinary week look like here. Then look for a genuine break - a sustained shift, not a single spike. Note when the shift started so it can be lined up against a known event.
How to read it
Two misreads to avoid. First, normal variance is not a trend: day-to-day wobble is expected, and one low day is noise, not a decline. Second, a level shift is not automatically a behaviour change - a new roster, a headcount change, a holiday week with fewer working days, or a coverage gap can move the line without anyone working differently. Always confirm the scope and the working-day basis are constant across the periods you are comparing before you call a change real. For utilisation, remember the denominator is available hours, so a trend can move because availability moved, not because activity did.
Honest limits
A trend shows direction, not cause. It cannot tell you why a line moved, and it does not prove the movement is good or bad - rising utilisation can mean recovered capacity or mean a thinner, more stretched team. It also says nothing about output: a flat activity trend can sit over rising or falling work volume. Confirm the explanation with the customer's own context and, for output, with Work Output.
Interpretation check
Separate signal from noise and from structure: a single spike is variance, and a level shift can come from roster, calendar or coverage changes rather than from how people worked. Hold scope and working-day basis constant before calling a trend real.
Practical task
Pick one team in the demo, read its baseline band over several weeks, and write one line stating whether the latest movement is noise, a structural shift, or a candidate real change - and what you would confirm to decide.
Self-check
Utilisation rises 6 points over a month. Why is that not automatically good news?
The denominator is available hours. The line can rise because availability fell (fewer people, more absence) rather than because activity grew. It also says nothing about output. Confirm what moved before celebrating.
A trend dips sharply for one day then recovers. How do you read it?
As normal variance, not a decline. One point is noise. Look for a sustained shift, not a single spike.
Two months are being compared but one had a public holiday week. What is the risk?
Different working-day counts move the line without any behaviour change. Hold the working-day basis and scope constant, or the comparison is misleading.
The line steps up the week a team was reorganised. What is the honest reading?
The scope changed, so this may be a structural shift, not a behaviour change. Note the event and confirm membership before attributing it to the team's effort.
Verified vs demo. Screen is live in the demo at /#/trends; exact controls and figures are SME-validation-pending.