# Workforce schedule exception case, step by step | SourceX

Canonical: https://refer.sourcex.si/guides/workforce-schedule-exceptions-exception-examples
Last updated: 2026-10-19

The workforce records that explain an exception are the ones that put the plan, the change, the reason and the result on one timeline: the published forecast and intraday reforecast, each schedule override with its approver and reason, the incident ticket that explains why volume moved, and service level by interval afterwards. This page walks through one fictional day, names who decided each step and which record it left, and ends with how to recognize a similar case at a client.

It is part of the [workforce schedule exceptions guide](/guides/workflow-workforce-schedule-exceptions), which covers what to inventory before a WFM renewal. Here the subject is a single hard case.

## The case: a firmware release floods the phones (Illustrative)

Fictional and invented for this page; not a real company, case or completed deal. A 410-person consumer electronics support operation runs voice and chat for its own products. Its service-level target is 80% of calls answered within 20 seconds. Volumes below are calls offered per half hour.

At 10:05 on an ordinary weekday, engineering releases a firmware update to a popular device. A fault in the update disconnects some devices from the home network. Customers call at once. The forecast, built weeks earlier from historical volume, expected about 125 calls per interval; by 10:30 the queue receives 365.

Nothing in the forecast was wrong by its own logic. WFM tools such as Zendesk WFM forecast from historical ticket data, and history held no record of this release. The exception is the gap between a sound plan and an event the plan could not see.

## Four decisions, interval by interval

Each row pairs what happened with the person who decided it and the record the decision left behind.

| Interval | Forecast | Actual | Service level | Decision and decider | Record it leaves |
|---|---|---|---|---|---|
| 09:30 | 120 | 118 | 82% | None needed | Published forecast, interval statistics |
| 10:00 | 120 | 140 | 74% | None yet; release goes out at 10:05 | Release ticket, first incident tickets |
| 10:30 | 125 | 365 | 31% | WFM analyst reforecasts; workforce manager cancels a training block for 14 agents at 10:20 | Reforecast version, schedule change request and approval |
| 11:00 | 125 | 390 | 44% | Head of customer operations moves 9 chat agents to voice at 10:40 | Skill reassignment, override note citing the incident ticket |
| 11:30 | 120 | 300 | 71% | Workforce director approves voluntary overtime at 11:05 | Overtime request, approval, adjusted time cards |
| 12:00 | 115 | 210 | 80% | Support lead adds a known-issue message to the phone menu | IVR change ticket |
| Next day | 120 | 160 | 81% | Analyst raises the week's forecast with a reason code | Forecast override with reason |

The reforecast at 10:30 is worth a note. Some platforms refresh intraday forecasts automatically; Amazon Connect documents a 15-minute refresh for queues with enough volume. Whether the analyst's version or the system's version was kept is a question to ask, not to assume.

## Why the override rows are the evidence

A staffing report would show the first four columns. The last two make this a record of judgment. Each override shows a person with authority choosing one action out of several under time pressure, citing a reason, and a measurable result one interval later.

Three features make the sequence useful as evidence rather than as a story someone remembers:

1. **The reason is linked, not paraphrased.** The 10:40 override cites the incident ticket by number. Anyone reading it years later can confirm why the move was made.
2. **The plan survives beside the change.** The original forecast and the reforecast are both kept, so the gap that triggered the decision is visible.
3. **The outcome is measured on the same clock.** Service level by half hour shows which action moved the number and how fast.

Adherence is another reason the published schedule matters. Amazon Connect calculates adherence as adherent time divided by adherent plus non-adherent time, against the published schedule. If the schedule in force at 10:20 was overwritten by the 10:40 version, the 14 agents pulled from training look non-adherent for the wrong reason.

Procurement exceptions follow the same pattern of request, approval and checked outcome; compare [how to assess procurement approvals and exception workflows](/guides/how-to-assess-procurement-approvals-and-exception-workflows) and [logistics exceptions and their documented resolutions](/guides/how-to-identify-logistics-exceptions-and-their-documented-resolutions).

## The same day, recorded badly

The contrast shows what to listen for. Here is how the same morning looks at an operation that kept less.

| What the strong record holds | What a weak record holds instead |
|---|---|
| Forecast and reforecast versions with publish times | Only the latest forecast; earlier versions overwritten |
| Override note citing the incident ticket number | Override note that says "spike" |
| Approver named on the schedule change | Approval given in team chat, never logged |
| Overtime request tied to adjusted time cards | Overtime visible only in payroll totals |
| Service level by half hour for the whole day | Daily service level average |

The weak version is not worthless; it still shows that something happened. But nobody can reconstruct who decided what, or whether it worked, so it explains far less about how the exception was handled.

## How to spot a similar case at a client

Ask the workforce director about the worst day of last year. Then listen for four things, all answerable without opening any record:

- **A named external cause.** A product release, an outage, a billing run or a weather event that the forecast could not have seen.
- **Several distinct actions.** Canceled training, channel moves, overtime, a phone-menu message. One action is a note; several are a sequence.
- **Reasons that point somewhere.** "We tie overrides to an incident ticket" or "we use reason codes" is the phrase that matters.
- **Kept versions.** Ask whether old forecasts and schedules are kept or replaced. The answer decides whether the plan can be read beside the change.

Then record metadata only: which systems, how many years, roughly how many exception days, who owns each record class. Agent names, adherence scores and recordings stay out of the conversation. The [records across systems map](/guides/workforce-schedule-exceptions-records-across-systems) shows how these records join, and [who can license workforce schedule exception records](/questions/who-can-license-workforce-schedule-exceptions-records) covers who to introduce. Recordings are a separate topic covered in [who owns call recordings in a contact center](/questions/who-owns-call-recordings-contact-center).

**Qualification.** The company needs 50+ full-time employees at peak (contractors excluded), a documented operating history, rights to license the records and an authorized contact. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 90% of the data is in English. Buyer interest, the rights result, the de-identification needed and any offer stay unknown until SourceX reviews the opportunity. Partners sign the Channel Partner Agreement before submitting introductions; the first valid introduction within the 90-day window is credited when the company submits a verified application. Other record types are on the [industries hub](/industries).

## Next step

> **Refer a company. Earn up to $100,000.** Default reward: 25% of Actual Supplier Payout, capped at $100,000 cumulatively per referred company. Eligibility and payment conditions apply, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.

If the workforce director can describe a day like this one with its overrides, reasons and interval outcomes still on file, [register as a partner](/register), check the company against [who qualifies](/who-qualifies) and read the [program terms](/terms) before you make the introduction.

## Sources

- https://refer.sourcex.si/terms
- https://docs.aws.amazon.com/connect/latest/adminguide/forecasting-capacity-planning-scheduling.html
- https://docs.aws.amazon.com/connect/latest/adminguide/schedule-adherence.html
- https://support.zendesk.com/hc/en-us/articles/6514633060634-Getting-started-with-Zendesk-WFM-Manager-guide