Real-Time Adherence: Intraday timeline
Objective. Read an agent's or team's state over the course of the day as a sequence, and use the timeline to explain a current adherence signal rather than judging a single moment in isolation.
The intraday timeline answers 'how did we get to the state we are in now?' - the chronological view that turns a single red cell into a story a lead can act on or dismiss with evidence.
See it in the demo
What this screen answers
The customer question is: across today, what sequence of planned and actual states led to where this agent or team is now? A Sales person uses it to show that RTA carries context, not just a live snapshot. A CSM uses it when justifying or dispositioning an exception - the timeline is where you confirm whether a current off-plan moment is a one-off, a pattern through the shift, or the tail of a legitimate reassignment that is already resolving.
The task
Scope to the agent or team and the current day, and read the planned track against the actual track along the time axis. Look for where the two diverge, how long each divergence lasted, and whether breaks, aux states and reassignments line up with what you expect. Use it to support an exception decision in the exceptions view and to brief the agent with specifics. Exact track labels, state colours and zoom controls are SME-validation-pending - drive at the level of 'read planned versus actual over the day'.
How to read it
The classic misread is judging an agent on the moment you happened to open the board. A single divergence can be a two-minute break overrun that grace rules should absorb, or a planned task swap - neither is a problem. The timeline's value is the shape over time: a repeated pattern of the same gap is a different conversation from one isolated event. Read duration and recurrence, not just presence of a red segment, and remember a long off-plan stretch may be a sanctioned reassignment, not drift.
Honest limits
The timeline shows the model on seeded demo data. Its accuracy depends on the schedule being the currently published version and on activity capture being complete for the period - a stale import or a capture gap will make the actual track misleading. Confirm with product how far back the timeline retains, its granularity and the data sources. It explains what happened; it does not prove why. Do not quote retention or latency figures as fact.
Interpretation check
Judge the shape over the day, not the moment you opened the board. A single red segment may be an absorbed break overrun or a sanctioned task swap; recurrence and duration are what distinguish drift from a legitimate one-off.
In the field
KnackGlobal (RCM): on a claims queue, a recurring same-interval gap through the shift is a staffing or process pattern worth fixing, whereas one reassignment segment is noise - the timeline is where a CSM tells them apart before escalating.
Practical task
On the demo, pick one agent with a visible divergence and write a 3-line read: when planned and actual diverged, how long and whether it recurs, and whether you would action it or dismiss it with the reason - citing duration and recurrence, not the single moment.
Self-check
You open the timeline and see one red segment mid-morning. Why is that not yet actionable?
One segment could be a brief break overrun grace rules should absorb or a planned swap. Check its duration and whether it recurs before treating it as drift.
What on the timeline would turn a signal into a genuine coaching conversation?
A repeated pattern of the same divergence through the shift, rather than a single isolated event - recurrence and duration, not one red cell.
Why can a timeline look wrong even when the agent did nothing unusual?
If the schedule shown is a stale import rather than the published version, or if activity capture has a gap, the planned or actual track is misleading. Validate the published-schedule version and capture coverage first.
Verified vs demo. The intraday timeline is live in the demo at /#/timeline on seeded data. Exact track labels, state colours, zoom controls, retention, granularity and data sources are SME-validation-pending; confirm with product.