Real-Time Adherence: Adherence rules
Objective. Understand the rules that govern what counts as adherent - grace windows, state mappings, thresholds, time zones - and why validating them is the precondition for trusting every other RTA screen.
Adherence rules answers 'what are the settings that decide whether an agent is adherent, and are they right for this customer?' - the configuration layer that every number on every other screen silently depends on.
See it in the demo
What this screen answers
The customer question is: what definitions turn raw activity and a schedule into an adherence verdict - how much grace before an overrun is a violation, how states map, what thresholds raise an exception, which time zone applies? A Sales person usually references this at awareness level to answer 'can you avoid flagging a 90-second overrun?' A CSM owns it: before any adherence number is trusted for an account, the rules behind it must be validated against how the customer actually works.
The task
As a CSM, walk the rules with the customer: confirm grace windows match their policy, state mappings reflect their real aux and work states, exception thresholds are set so the queue is useful rather than noisy, and the time zone and published-schedule source are correct. This is a validation and tuning task, often with specialist escalation for advanced setup, not a casual toggle. Exact setting names, default values and the permissions to change them are SME-validation-pending - describe what to validate, not specific fields.
How to read it
The key realisation is that adherence is not an absolute truth - it is the output of these rules. Change the grace window and yesterday's violations become today's non-events. So the misread to avoid is debating an adherence number without first checking the rules that produced it. If a customer disputes a figure, the rules screen is the first stop: wrong time zone, a stale schedule source or an unmapped state will make every downstream screen wrong in a way no amount of floor review can fix.
Honest limits
This screen shows the configuration model on seeded demo data. The specific rules available, their defaults, who may edit them and how they interact are production and entitlement dependent - confirm with product rather than asserting a setting exists. Correct rules make the data trustworthy but do not create data: if activity capture or the schedule feed is missing, no rule fixes that. Treat advanced configuration as a specialist-escalation area and do not promise a particular rule capability as fact.
Interpretation check
Adherence is the output of these rules, not an absolute truth. Never debate or trust an adherence figure without first validating the grace windows, state mappings, thresholds, time zone and schedule source that produced it.
In the field
KnackGlobal (RCM, multi-site): time zone and published-schedule-version settings are exactly where a multi-site operation gets adherence wrong, so validating the rules first is what keeps a 6,000-seat rollout from arguing about numbers that were mis-defined.
Practical task
On the demo, review the adherence rules and write a 4-item validation checklist you would run with a real customer before trusting their adherence numbers: grace windows, state mappings, exception thresholds, and time zone plus published-schedule source - noting which items you would escalate to a specialist.
Self-check
A customer disputes an adherence number. Where do you look first and why?
The adherence rules. A wrong time zone, stale schedule source or unmapped state will make every downstream screen wrong at the source - no floor review fixes a definition error. Validate the rules before the figure.
Why is 'adherence' not an absolute measurement?
It is computed from configurable rules - grace windows, state mappings and thresholds. Change the grace window and the same behaviour flips between violation and non-event. The rules define what adherent means.
The exception queue is too noisy. Is that a floor problem or a rules problem first?
Check the rules first. Thresholds set too tight raise exceptions on trivial overruns. Tune the grace and thresholds before concluding the floor is poorly run.
Verified vs demo. Adherence rules is live in the demo at /#/settings on seeded data. Exact setting names, default values, edit permissions and advanced configuration options are SME-validation-pending; confirm with product and treat advanced setup as specialist escalation.