WFM guide

Real-time adherence (RTA)

Q: What is real-time adherence (RTA) in a contact centre?

Real-time adherence (RTA) is a live comparison between what an agent is scheduled to be doing at this moment (from the WFM schedule) and what the agent is actually doing (from the ACD or telephony system). The comparison is made at interval level, typically every 15 seconds to 1 minute. If the scheduled state (e.g. 'available') does not match the actual state (e.g. 'extended break'), an exception is triggered. RTA enables WFM analysts and team leaders to identify and respond to coverage shortfalls in real time, rather than discovering them in the following day's adherence report.

Q: What is the difference between real-time adherence and schedule adherence?

Real-time adherence (RTA) is a live monitoring function — it shows the current deviation between schedule and actual state for each agent right now. Schedule adherence is a historical reporting metric — it shows the percentage of scheduled time that agents were in the correct state during a past period (shift, week, month). RTA informs intraday response decisions ('this agent has been off-schedule for 12 minutes — do we need to reallocate?'). Schedule adherence informs performance management decisions ('this agent's monthly adherence is 78% — this needs a coaching conversation').

Q: What triggers an RTA exception?

An RTA exception is triggered when the agent's current ACD state does not match their scheduled state and the deviation exceeds a threshold. Common exceptions: (1) Extended break, where the agent is in break state beyond scheduled break end time (threshold typically 3–7 minutes). (2) Unplanned absence, where the agent is not logged in at the start of a scheduled shift or returns late from lunch (threshold typically 5–10 minutes). (3) Unplanned off-phone, where the agent is in a non-customer-facing state (training, meeting) without a planned schedule entry. (4) Unplanned overtime, where the agent is still logged in beyond scheduled shift end. (5) Wrong state, where the agent is in after-call work when they should be available, or vice versa.

Q: What should WFM analysts do when an RTA exception fires?

RTA exception response depends on severity and queue state. For minor deviations (under 5 minutes, queue stable): log the exception and allow natural resolution — intervening on every minor deviation creates a surveillance culture and trains agents to be resentful rather than responsive. For moderate deviations (5–15 minutes, queue building): alert the team leader so they can address with the agent. For major deviations (over 15 minutes, or concurrent with a queue critical state): escalate immediately; consider queue recovery actions (accelerate break return, request overtime, initiate overflow to alternate skill group). The WFM analyst's role is to triage and escalate, not to manage individual agent behaviour directly; that is the team leader's responsibility.

Real-time adherence monitoring compares every agent’s current ACD state against their schedule, second by second. The goal is not surveillance. It is early warning: identifying coverage shortfalls while there is still time to act, rather than reading about them in tomorrow’s report.

How RTA monitoring works

The RTA comparison loop

WFM schedule

Agent scheduled state at this moment: e.g. 'Available on phones 09:00–11:30'

→

ACD live state

What the ACD sees now: e.g. agent in 'After-call work' at 10:42

→

Compare

States match? No exception. States differ by more than threshold? Exception fires.

→

Alert

WFM dashboard highlights agent; team leader or WFM analyst reviews and acts

Most WFM platforms poll the ACD for agent state every 15–60 seconds. The threshold (how long a deviation must persist before an exception fires) is configurable, typically 3–7 minutes for break exceptions and 5–10 minutes for login exceptions.

RTA vs. schedule adherence reporting

Real-time adherence (RTA)

Timeframe: Now, live, current moment
Purpose: Intraday action and queue protection
Who uses it: WFM analyst (triage and escalate), team leader (act)
Typical response: Minute-to-minute; alert, investigate, adjust
Output: Active exception list with agent and duration

Schedule adherence reporting

Timeframe: Past, by shift, week, or month
Purpose: Performance management and trend analysis
Who uses it: Team leader, operations manager, HR
Typical response: Weekly or monthly; coaching, formal review
Output: % adherence per agent per period

Common confusion: RTA and adherence reporting are not the same metric. An operation can have excellent end-of-day adherence (agents return from overlong breaks and make up the time) while experiencing severe intraday SL failures (the queue collapsed during the overlong break). RTA protects intraday service level; adherence reporting measures a different outcome.

RTA exception types and thresholds

Thresholds are guidelines; calibrate to your operation size and queue sensitivity.

ExceptionAmberRedResponse

Extended break

Agent remains in break state beyond scheduled break end time

3–5 min overrun7+ min overrunTeam leader action; WFM notes for end-of-day adherence report

Late login / late return

Agent not logged into ACD at scheduled shift start or late returning from lunch

3–5 min late10+ min lateTeam leader confirms agent is in and delay reason; WFM flags for attendance record

Extended ACW (after-call work)

Agent in after-call work state significantly beyond normal ACW duration for contact type

2× typical ACW4× typical ACWInvestigate whether it is a legitimate complex case, a system issue, or avoidance behaviour; each requires different action

Unplanned off-phone

Agent in non-customer-facing state (meeting, coaching, admin) with no approved schedule activity

5 min unplanned15+ min unplannedVerify whether agent was pulled off-phone by their team leader without WFM notification, which is common in informal operations

Schedule state mismatch

Agent in a different approved state than scheduled (e.g. in training during scheduled phone time)

Any mismatch during peak intervalAny mismatch during red-queue intervalAssess queue impact; adjust if needed; update schedule to reflect actual activity

Prolonged available state (ghost available)

Agent in 'available' state but not receiving contacts, indicating a potential ACD routing or technical issue

5+ min available with no contact in a high-volume period10+ min available with queue depth buildingCheck ACD routing configuration; verify agent skill group assignment; may indicate system fault

What WFM should and should not act on

Over-intervention on RTA exceptions creates surveillance culture and damages agent engagement without improving service level. Under-intervention allows queue failures to compound. The distinction matters:

WFM analyst: triage and escalate

→ Monitor exception list and queue state simultaneously
→ Prioritise exceptions that coincide with queue degradation (SL dropping, ASA rising)
→ Escalate to team leader for individual agent action; do not contact agents directly
→ Consider queue recovery actions for red-level exceptions (overflow, overtime request)
→ Log exceptions for end-of-day adherence reporting regardless of intraday action taken

What WFM should not do

→ Contact agents directly when exceptions are minor and queue is stable
→ Use RTA data to build informal individual performance cases; that is adherence reporting's role
→ Alert on every deviation regardless of threshold or queue state (alert fatigue degrades response quality)
→ Make staffing decisions based on a single interval exception; look for patterns, not single events
→ Assume every extended ACW is avoidance; some is legitimate, complex post-call work

RTA, surveillance, and agent trust

RTA is one of the most sensitive topics in contact centre management. Used well, it protects queue and gives WFM the visibility to fix coverage problems before they become SL failures. Used poorly, it becomes a tool for micromanagement that damages trust and drives disengagement.

Be transparent about what is monitored

Agents should know that their ACD state is compared to their schedule in real time. Covert monitoring creates resentment when agents discover it. Transparent monitoring, where agents understand what triggers an alert and why, is better received and produces fewer evasion behaviours.

Explain the queue rationale, not the attendance rule

When team leaders act on an RTA exception, the explanation should be operational ('the queue has 40 people waiting and we need you back') not disciplinary ('you are 7 minutes over your break'). The former creates shared ownership of the outcome; the latter creates compliance without understanding.

Reserve formal action for patterns, not single events

A single 8-minute overrun break triggers the RTA alert and deserves acknowledgement. It does not deserve a performance warning. Formal performance action should be based on persistent adherence patterns from reporting data, not individual intraday exceptions.

Real-time adherence questions

What is real-time adherence (RTA) in a contact centre?

RTA is a live comparison between the agent's scheduled state and their current ACD state, made every 15–60 seconds. When the states differ by more than a configured threshold, an exception fires. The goal is intraday visibility: identifying coverage gaps while there is still time to act, rather than discovering them in the next day's report.

What is the difference between real-time adherence and schedule adherence?

RTA is live monitoring for intraday action: it shows current deviations and informs the response right now. Schedule adherence is a historical metric that shows the percentage of scheduled time agents were in the correct state during a past period. RTA informs intraday decisions; adherence reporting informs performance management.

What triggers an RTA exception?

An exception fires when the agent's ACD state does not match their scheduled state and the deviation exceeds a threshold. Common triggers: extended break (3–7 min threshold), late login (5–10 min), prolonged ACW, unplanned off-phone, scheduled state mismatch, or ghost-available (in 'available' but not receiving contacts).

What should WFM analysts do when an RTA exception fires?

Triage by severity and queue state. Minor deviation, queue stable: log and allow natural resolution. Moderate deviation, queue building: alert the team leader. Major deviation or red queue state: escalate immediately, consider queue recovery actions. WFM analysts should escalate to team leaders rather than contacting agents directly.