Scheduling: Optimization
Objective. Run optimization on a draft and compare shortage, overtime and violations before and after, noting which rules were respected - without treating the result as published or as a fix for a wrong forecast.
Optimization answers: can this draft be reshaped to cut shortage and overtime while respecting rules, cost and employee constraints? The critical read is that it reshapes the draft against the inputs you gave it - it does not change the forecast, and its output is still a draft.
See it in the demo
What this screen answers
A Sales or CSM person opens optimization to answer: can the draft be improved against rules and cost - reducing the shortage window and overtime while respecting skills and labour rules? In the three-step story this is step two, and the whole point is that it is a distinct step operating on the draft from the planner.
The task
On the W42 draft with its known shortage window, run optimization and compare the before and after: does the shortage shrink, does overtime fall, do rule violations clear - and which rules and skills were respected while doing so? Capture the before/after as the heart of the demo narrative.
How to read it
Optimization reshapes the draft against the rules, cost and constraints you set. It does not change the forecast underneath it - the improvement is relative to the inputs you gave it. Garbage demand in, confident-but-wrong schedule out. A better result is not proof the week is handled: the output is still a draft, not a live schedule, until it is published. State this explicitly: 'optimization reshapes the draft, it does not fix the forecast.'
Honest limits
Optimization cannot repair a wrong forecast, a missing labour rule, or an incorrect skill map - it will optimise confidently against whatever it is given. It is a step, not magic. Confirm with product which labour rules, cost parameters and constraints the optimiser actually enforces for a customer; the demo shows the model on seeded data.
Interpretation check
Optimization improves a draft against the rules and forecast you set; it cannot fix a wrong forecast or a missing rule, and its output is still a draft, not a live schedule.
In the field
An RCM back office cutting weekend overtime: optimization reshapes the draft to respect rest rules and cut premium cost - but only after the CSM confirms the forecast it is built on is the live one.
Practical task
On the demo, run optimization on one shortage window and record the before/after for shortage, overtime and violations, plus one sentence on what optimization did NOT change (the forecast).
Self-check
A customer says optimization will 'fix everything automatically.' What is the honest framing?
It improves a draft against rules and cost. It cannot repair a wrong forecast or missing rules - correct the inputs first, then optimise. It is a step, not magic.
After a strong optimization result, is the week covered?
No. The result is still a draft. It becomes a commitment only when it is published, on the correct forecast and with violations cleared.
Why insist optimization is a distinct step from planning and publication?
Because conflating them is the biggest misread: a draft is not live, and optimization reshapes the draft against rules/cost but never changes the forecast beneath it.
Verified vs demo. Screen is live in the demo at /#/optimize; exact controls, labels and figures are SME-validation-pending.