A Schedule Is a Forecast
Every published rota is a bet about next week's demand. Treating it as a plan rather than a prediction hides the thing worth measuring.
A schedule says how many people will be where, when. That is a claim about the future, and like every claim about the future it has an accuracy that can be measured.
The planning problem in “A Schedule Is a Forecast” becomes clearer when scheduled hours can be compared with actual project and time records. Teams researching this product overview for daily schedule template can add that operational evidence, while demand, service levels and manager judgement remain necessary to explain why a variance occurred.
What the bet consists of
That demand will be roughly this shape.
For an independent reference related to “A Schedule Is a Forecast”, consult the CIPD workforce-planning resources; it provides a useful external check on scheduling, working-time and workforce-planning assumptions.
That these people will be available.
That the work will take about as long as it usually does.
Three predictions, compounding, and the schedule is right only if all three hold.
Why calling it a plan is the problem
A plan is something you execute.
A forecast is something you check afterwards and improve.
Operations that treat the schedule as a plan never ask whether it was right — they ask whether it was followed, which is a different and less useful question.
What measuring accuracy requires
Planned hours against demand actually experienced, by period.
Not planned against actual hours, which measures compliance.
The gap between forecast demand and real demand is the thing that drives everything else, and most operations do not record forecast demand at all.
The simplest version
Write down what you expected, before the week.
One figure per day, or per shift if the shape matters.
Compare afterwards. Keep the pairs.
After eight weeks you know your error, which is more than almost any operation knows.
What the pairs tell you
The size of your typical error.
Whether it has a direction — most forecasts are biased, consistently.
Which days or periods are hardest to predict.
And whether a particular event or condition throws it, which is the next note.
Why direction matters most
A forecast that is wrong by the same amount in both directions needs more capacity to absorb variation.
A forecast that is consistently low needs correcting, and the correction is arithmetic.
Most operations have a bias and have never looked for it, which is a free improvement.
The cultural piece
If a wrong forecast is treated as a mistake, people stop recording what they expected.
Which removes the only route to improving it.
Record the expectation, compare it, and treat a large error as information about the week rather than about the person.
What to check
Do you write down expected demand before the week?
Could you say how wrong you usually are?
Does your error have a direction?
And is forecast accuracy reported anywhere, or only hours against budget?