Intraday Management assessment - prove you can run the loop and hold the honesty line
Objective. Test mastery of IDM for the certification gate: can the learner run the detect - decide - act - review loop on a real shortage, read every figure with the correct boundary, and never present a projection as a result. Scored to the gate - practical score at least 80, no critical error, manager review.
This assessment puts the learner on a live-ish IDM shortage and a customer dispute and scores how they decide and how they talk. The pass bar is not whether they can click the screens - it is whether they keep simulated, executed and achieved apart, read coverage by interval and skill, check the rule behind a disputed figure, and hold the execution boundary out loud. Any collapse of projection into result is a critical, failing error.
See it in the demo
Scenario 1 - run the loop on a real shortage (CSM-led, both roles)
Seeded setup: by 11am the control tower flags a skill group running off plan. Forecast vs actual shows a gap that is widening, not recovering. Staffing shows the shift roughly balanced overall but one skill short in the 1pm-3pm intervals with a surplus elsewhere.
The learner must: confirm the drift is real and sustained (not act on the tower flag alone); read the gap by interval and skill rather than the net total; shortlist two corrective actions that fit the specific gap (for example a skill move from the surplus, or overtime); simulate both in the what-if and compare projected coverage, backlog and cost; choose one with a stated rationale; log the decision with its projection and owner; and name the end-of-day review they would run to check it. Throughout, every projected number must be labelled a projection, and the execution boundary (change executes in the customer's systems) stated.
Scenario 2 - the disputed figure and the 'does it auto-fix?' prospect (sales-led, both roles)
Seeded setup: a prospect in a live demo says two things. First: 'Your what-if said opening overtime gives us +5 percent coverage - so that is what we would get, right?' Second: 'And it just runs these fixes for us automatically?'
The learner must answer both honestly on the spot: correct the first by separating simulated from executed from achieved - the +5 percent is a projection on current assumptions, to be executed in the prospect's systems and then checked against actual; correct the second by stating the execution boundary plainly - IDM detects, reforecasts, simulates, recommends and logs, execution happens in the prospect's systems within policy, confirm with product exactly what it can trigger versus hand off. A bonus-credit learner volunteers these boundaries before being pushed and reframes to the honest value: a faster, evidenced, auditable decision, with the service outcome baselined not promised.
Scored checklist - what a correct answer contains
Scenario 1 (max 50):
- Confirms drift on forecast vs actual before acting, not on the tower flag alone (10)
- Reads the gap by interval and skill, explicitly rejecting the net-balance reading (10)
- Simulates and compares two fitting actions rather than betting on one (10)
- Logs the decision with projection and owner, and names the end-of-day review (10)
- Labels every projected figure as a projection and states the execution boundary (10)
Scenario 2 (max 50):
- Correctly separates simulated, executed and achieved on the +5 percent (15)
- States the execution boundary plainly and defers execution specifics to 'confirm with product' (15)
- Reframes to the honest value - decision quality and speed, service outcome baselined not promised (10)
- Volunteers the boundaries proactively / professional, credible delivery (10)
Pass: total at least 80 of 100, no critical error, and manager review of the recorded response.
Critical errors that fail the assessment regardless of score
Any one of these fails the learner even if the numeric score is high, because each would damage a real customer relationship or a deal:
- Presenting a what-if projection as a delivered or guaranteed result (the core IDM failure).
- Claiming IDM executes the fix / runs the day automatically, or being unable to state the execution boundary.
- Reading a net staffing balance as 'the day is covered' and missing the interval/skill shortage.
- Acting on a tower flag or an alert as a verdict without confirming the drift is real.
- Attributing the service / backlog outcome to IDM rather than to the customer's own execution and baseline.
- Asserting on entitlement, cost-model internals, feed latency or what IDM can trigger directly, instead of escalating or 'confirm with product'.
These map one-to-one to the signature guardrail: variance is a prompt not a verdict, a reforecast is a decision aid the user approves, 'on plan' at noon is not an end-of-day guarantee, and the engine proposes while the human disposes.
Interpretation check
The assessment passes learners who keep simulated, executed and achieved apart, read coverage by interval and skill, check the rule behind a disputed figure, and hold the execution boundary out loud. It fails anyone who presents a projection as a result or claims IDM runs the day - the engine proposes, the human disposes, and the service outcome is the customer's to baseline.
In the field
KnackGlobal (RCM, ~6,000 seats) and OVO (contact centre) are the anchor contexts for the scenarios - a time-sensitive claim queue and a volatile contact-centre afternoon respectively. Both are places where presenting a projection as a result, or claiming automated execution, would damage a real, strategic relationship - which is why those are the assessment's critical errors.
Practical task
Sit the assessment on the seeded demo account. Record yourself working Scenario 1 end to end (detect, confirm, locate by interval and skill, simulate and compare two actions, choose, log with projection and owner, name the end-of-day review) and answering Scenario 2's two prospect questions live. Self-score against the checklist, flag any critical error honestly, and submit the recording plus your score for manager review. You pass at 80 or above with zero critical errors and a passed manager review.
Self-check
ASSESSMENT ITEM (Scenario 1): The tower flags a skill group red at 11am. A learner immediately opens overtime. What is wrong, and what should they have done first?
MODEL ANSWER: Acting on the tower flag as a verdict is a critical error - it is a prompt to investigate, not an instruction to act. They should first confirm on forecast vs actual that the drift is real and sustained (widening, not a recovering blip), then locate the gap by interval and skill on staffing, then simulate and compare two fitting actions in the what-if before deciding. Acting blind on a flag fails the assessment.
ASSESSMENT ITEM (Scenario 1): Staffing shows total required equals total available. The learner concludes the day is covered. Score this.
MODEL ANSWER: Critical error - reading a net balance as coverage. A balanced total can hide a severe interval or skill shortage next to an idle surplus, which is exactly what breaks service. The correct answer reads the gap by interval and by skill, identifies the 1pm-3pm skill shortage, and notes the surplus as raw material for a skill move. This failure alone fails the assessment regardless of other points.
ASSESSMENT ITEM (Scenario 2): The prospect says 'your what-if said +5 percent coverage, so that is what we would get'. Give the passing answer.
MODEL ANSWER: 'That +5 percent is a projection on current assumptions - it is what the model expects if arrivals and handling behave as assumed and the action is actually executed. A simulated improvement is not an executed action, and an executed action is not an achieved service result. We would execute it in your systems and then check it at end of day against actual. The honest value is a faster, evidenced decision - the service number we would baseline with you, not promise.' Separating simulated / executed / achieved is required to pass; presenting +5 percent as delivered is a critical error.
ASSESSMENT ITEM (Scenario 2): The prospect asks 'does it just run these fixes automatically?' Give the passing answer and name the trap.
MODEL ANSWER: 'No - IDM detects drift, reforecasts, simulates options, recommends and logs the decision. The actual staffing change executes in your own systems within your cost and policy rules. Confirm with product exactly what it can trigger versus hand off.' The trap is claiming automation IDM does not perform, or being unable to state the boundary - either is a critical, failing error. Bonus credit for having volunteered this before being asked.
ASSESSMENT ITEM: A learner scores 88 on the checklist but attributed the day finishing on target to IDM. Do they pass?
MODEL ANSWER: No. Attributing the service outcome to IDM is a listed critical error and fails the learner regardless of the numeric score - the outcome is the customer's to own and baseline, and many factors move a day. IDM evidences the decision's quality and speed; it does not deliver the service number. The gate requires at least 80, no critical error, and manager review - the critical error overrides the score.
Verified vs demo. Scenarios run on seeded demo data at idm.demo.timechamp.io; all figures are illustrative and SME-validation-pending. The assessment tests judgement and honesty, not demo numbers. What IDM executes directly versus hands off, entitlement and feed latency are production-dependent - a correct answer defers these to 'confirm with product' rather than asserting them, and doing so is scored as correct, not evasive.