Genesys Cloud Interaction Model Timeline
An interactive exploration of how Genesys Cloud stacks three live status systems on the WFM schedule, and how a single agent holds several interactions at once across voice, chat, and email.
Genesys Cloud · WFM
One agent, four status systems, and the slots they fill
Genesys stacks three live status systems on the WFM schedule, then lets a single agent hold several interactions at once - up to a capacity set per media type, not one overall limit. Scrub the cursor and watch primary status, routing status, the real contacts, and per-channel load line up. The adherence and concurrency lanes aren't drawn by hand; they're computed minute-by-minute from the lanes above.
Reading the lanes
Scheduled is the WFM plan as activity codes. Primary status (presence) is what the agent showed. Routing is system-set and only lives while On Queue. Interactions stack into slots - the agent works more than one at a time. Concurrency counts live interactions. Adherence is the verdict.
Capacity is configured per media type - Genesys defaults to voice 1, email 1, chat 4 (each adjustable up to 25). There's no single overall cap: when one channel hits its max, routing stops offering that channel, but the agent can still be sent a different one. A slot stays held through after-call work, not just the talk.
This shift · Monday
09:00 - late on queue. Scheduled On Queue Work, but the agent sat in Available for six minutes. Available maps to Off Queue, so adherence counts it out.
11:13 - the phantom violation. On Queue, scheduled right, nothing visibly wrong. But a missed offer dropped routing to Not Responding, so the ACD benched them and adherence goes red.
13:28 - early back from meal. Returning two minutes early is also out of adherence - the actual activity no longer matches the scheduled one. WFM scores the match, not the effort.
15:06 - blended load. Two web chats and an email at once. Chat sits at 2 of 4 so routing could send another chat; email is at 1 of 1, so no second email comes.
A mis-mapped secondary status manufactures the same phantom red: the schedule expects one activity code, the reason code under the status points elsewhere, and they never meet.
Metric families & how Genesys classifies time
The three live lanes at the top of the timeline are the three status families. Each produces its own time metrics, and WFM rolls those up into the productive / shrinkage picture. There's no single switch labelled "productive" - it's built from activity-code attributes (paid, work time, shrinkage) plus where the routing status lands.
The presence the agent sets, or the system sets. One at a time. Decides whether incoming calls go to voicemail.
Available · Busy · Away · Break · Meal · Meeting · Training · Out of Office · On Queue · Offline
Metrics: time in status · On-Queue time · Off-Queue time · logged-in time
Org-defined reason code under a primary status. This is what adherence maps to activity codes.
e.g. Away · Restroom · Busy · Coaching · Break · Paid
Metrics: time in each secondary status, live and historical (incl. busy secondary statuses)
System-set, only while On Queue. Says whether the ACD counts the agent as working a contact.
Off Queue · Idle · Interacting · Communicating · Not Responding
Metrics: Interacting · Idle · Communicating · Not-Responding time
How each kind of time is classified
| Activity / status | Category | Paid | Work time | Productive? | Shrinkage |
|---|---|---|---|---|---|
| On Queue · Interacting | On Queue Work | paid | yes | Productive | none |
| On Queue · Idle (ready) | On Queue Work | paid | yes | Productive | none |
| After-call work (ACW) | On Queue Work | paid | yes | Productive | none |
| Not Responding (on queue) | — | paid | no | Non-productive | unplanned |
| Break | Break | paid | no | Non-productive | paid · planned |
| Meal | Meal | unpaid | no | Non-productive | unpaid · planned |
| Meeting | Meeting | paid | no | Non-productive | paid · planned |
| Training | Training | paid | no | Non-productive | paid · planned |
| Time off · PTO | Time Off | paid | no | Non-productive | paid · planned |
| Time off · unpaid leave | Time Off | unpaid | no | Non-productive | unpaid · planned |
| Sick (called in) | Time Off | paid / unpaid | no | Non-productive | unplanned |
| Away / Busy off queue (unscheduled) | Unavailable | usually unpaid | no | Non-productive | unplanned |
Paid and shrinkage are per-activity-code settings, so a break may be paid in one org and a meal unpaid in another - the rows show common defaults, not fixed rules. The eight categories are fixed; the codes inside them are yours to configure.
The rolled-up WFM metrics
Productive time
On-Queue time the agent spends handling or ready to handle: Interacting + ACW + Idle. The activity-code work-time flag is true. This is capacity actually applied to, or available for, contacts.
Non-productive time
Everything that isn't work time - breaks, meals, meetings, training, away/busy, not-responding. Paid or unpaid, scheduled or not. All of it is shrinkage.
Paid shrinkage
Scheduled paid time the agent isn't available for interactions: paid breaks, meetings, training, paid PTO. You pay for it and handle no contacts during it - the expensive kind.
Unpaid shrinkage
Non-available time that isn't paid: unpaid meals, unpaid leave. Reduces coverage without payroll cost, but still has to be staffed around.
Shrinkage %
Share of scheduled time lost to non-interacting activity - planned (baked into the shift) plus unplanned (sick, no-show, non-adherence). Drives how many extra agents to schedule.
Adherence vs Conformance
Adherence asks whether actual status matched the scheduled activity at each interval (the verdict lane). Conformance asks only whether the agent worked the right total on-queue time, ignoring exact timing.
Occupancy
Of on-queue time, the slice spent Interacting versus Idle. High occupancy means little idle headroom; it's a productivity-pressure gauge, not an adherence one.
Where each comes from
Status times come from the analytics / agent-status views; adherence, conformance and shrinkage come from WFM using the schedule and the activity-code flags. Same raw statuses, different roll-ups.
Illustrative sample shifts, not live data. Routing statuses, the eight WFM activity categories, agent-utilisation defaults, and adherence rules follow Genesys Cloud documentation at help.genesys.cloud. Adherence and concurrency are computed in-browser from the data, not hard-coded.
Work with us
Want to explore Genesys Cloud WFM for your operations?
Book a 15-minute working session to discuss WFM selection, WFM configuration, and keeping humans in control in an agentic AI world.
Book a working session