Read metrics with context
Objective. Before interpreting any Time Champ number, check its denominator, timezone, exclusions and freshness, and know why attendance and utilisation answer different questions - so you never make a claim or a decision on a number you have not qualified.
This is the single most important analytical habit in the programme. It teaches the five questions to ask of every metric before you trust it, the attendance-versus-utilisation distinction that trips up everyone, and how both sales and CSM use this to avoid the credibility-destroying mistake of acting on an unqualified number.
See it in the demo
The customer problem
Dashboards are persuasive. A percentage in a large font feels like truth, and people act on it - escalate a team, question an employee, size a business case - before asking what it actually measures. But the same screen can show two very different realities depending on its denominator, the timezone it is rendered in, what it excludes, and how fresh the data is. A utilisation figure that looks alarming may be dividing by the wrong hours; an adherence number may be stale; a comparison may span two timezones. The business impact is severe and specific: managers escalate the wrong things and lose trust in the tool, executives fund decisions on false precision, and a single public mistake with a bad number can poison adoption for a whole team. The skill that prevents all of this is cheap - a handful of questions asked every time before interpretation.
The five context questions
Ask these of every metric, every time, before you say what it means:
- Denominator - what is this divided by? Utilisation over available hours is a different number from utilisation over paid hours or scheduled hours. If you do not know the denominator, you do not know the metric.
- Timezone - whose clock? A cross-site comparison or an adherence view in the wrong timezone silently misaligns intervals. The Workforce Management apps are especially sensitive here.
- Exclusions - who and what is in or out? Are leave, breaks, non-counted work types, or decommissioned clients included? Exclusions change the story more than most people expect.
- Freshness - how current is this? A simulated or not-yet-refreshed number is not today's reality (Intraday's rule: a simulated improvement is not an executed action; Scheduling's rule: a draft on an old forecast can look valid while failing the current requirement).
- Source and definition - where did it come from and how is 'done' or 'productive' defined? Coverage gaps and definition drift are the usual root cause of a mismatch.
Only after these five do you interpret.
Attendance vs utilisation (and other confusable pairs)
The canonical trap: attendance and utilisation are not the same question. Attendance asks 'was the person present and for how long?' - it reconciles paid/available presence. Utilisation asks 'of the available time, how much was used productively?' A team can have excellent attendance and poor utilisation, or vice versa; quoting one as if it answered the other is a classic error (Time and Activity's own interpretation rule).
The same discipline applies across the platform:
- Schedule-adherent is not output-adequate - an agent can follow the schedule and be active yet still be below the output standard (RTA's rule).
- A non-productive label is not proof of no useful work (Employee Monitoring's rule).
- High evaluation coverage is not scoring accuracy (Quality's rule).
- An overall FTE balance is not per-skill, per-interval coverage (Capacity Planning's rule).
Each pair looks similar and answers a different question. Knowing which question a number answers is most of the job.
What to say / do
Sales, in a demo: when you show a number, qualify it out loud - 'this utilisation is over available hours, for this team, this week, excluding leave.' Doing this builds more trust than a bigger number would, because it signals you will not sell them false precision (ties to S04's required explicit limitation).
CSM, investigating a discrepancy (the C05 routine): run the five questions as a checklist before concluding. Most 'the numbers are wrong' tickets resolve at denominator, exclusions or freshness, not at a product bug. Write the reconciliation up with the context you checked, so the customer learns the habit too.
Both roles: when a customer quotes a number back at you to make a decision, your first move is not to agree or disagree but to ask 'what is that over, and how fresh is it?' Model the habit and it becomes the account's habit.
Measurement and honest limits
Freshness and definition are themselves measurable: data coverage, source freshness and metric-definition clarity are candidate health signals a CSM tracks (they recur across the app grounding). A metrics programme that cannot state its own denominators and freshness is not ready to drive decisions.
Honest limits: the demos show the metrics and the interpretation rules, and they are built on seeded demo data. They do not prove the denominators, exclusions, timezones and freshness of a specific customer's production instance - those must be validated in that instance. Never quote a demo number as if it were a customer result, and never let a customer treat a number as decision-ready until its five context questions are answered. When a mismatch cannot be reconciled with the five questions, escalate with a clear packet (what you checked, what remains) rather than guessing (C05, F09).
Interpretation check
Attendance and utilisation answer different questions (presence versus how available time was used); the same is true of schedule-adherent versus output-adequate, and of a non-productive label versus proof of no work. Always know which question the number answers before interpreting it.
In the field
In a KnackGlobal displacement comparison, reading with context is what keeps the story honest: output-per-FTE is only comparable once denominator (which paid hours), exclusions (work type, tenure) and period are aligned across the teams being compared - exactly the Work Output rule that two teams are compared only after aligning work type, complexity, period, tenure and quality.
Practical task
Take one metric from the Time and Activity demo and one from Real-Time Adherence and write a short 'read with context' note for each: state the denominator, timezone, exclusions, freshness and definition as best the demo shows, flag what you would have to verify in a real instance, and write the one sentence you would say to a customer before letting them act on it.
Self-check
A team shows 98% attendance and 61% utilisation. What does that combination actually tell you, and what does it not?
People are reliably present but a large share of available time is not being used productively; it tells you where to investigate available-time use. It does not by itself prove under-performance - you must still check the utilisation denominator, exclusions and what 'productive' means.
Why is 'the schedule shows full adherence' not the same as 'the work got done'?
An employee can be schedule-adherent and active yet still below the output standard; adherence measures following the plan, not producing the required output. They are separate adherence layers (RTA).
List the five context questions and name the one that most often resolves a 'numbers are wrong' ticket.
Denominator, timezone, exclusions, freshness, source/definition. In practice, denominator, exclusions or freshness resolve most disputes before any product bug is suspected.
A customer quotes a dashboard number to justify cutting a team. What is your first response?
Not agreement or disagreement, but 'what is that number over, and how fresh is it?' - qualifying the denominator and freshness before any decision rests on it.
Verified vs demo. The metrics and interpretation rules are shown on seeded demo data; the actual denominators, exclusions, timezones and freshness of a customer's production instance must be validated there and never inferred from the demo.