Forecasting
Seasonal Index & Forecast Calculator
Split a demand history into the level and the shape of the year. A seasonal index says what each part of the cycle is worth against an ordinary one: 1.47 means that quarter runs 47% above average and always has. With the shape separated out you can forecast the year as one number and let the indices put it back into months, which is how almost every real demand plan is actually built.
Inputs
- Seasons in a cycle — Quarters is 4, months is 12. A cycle is whatever repeats, which is almost always a year.
- Cycles of history — Two years of history is the fewest that can tell a season from a one-off. One is accepted and the result says what it is worth.
- Cycle 1, season 1 — The oldest cycle first, and the seasons in calendar order inside it.
- Cycle 1, season 2
- Cycle 1, season 3
- Cycle 1, season 4
- Cycle 2, season 1
- Cycle 2, season 2
- Cycle 2, season 3
- Cycle 2, season 4
- Total demand you expect next cycle — Your own forecast for the whole of next year. The indices decide how it is split, not how big it is.
How to use this calculator
- Set how many seasons a cycle has — 4 for quarters, 12 for months — and how many cycles of history you are entering. The demand boxes follow both numbers, oldest cycle first.
- Enter what you actually sold in each season, not what you forecast or what you shipped. A stockout month looks like a weak season to this calculation, and it was not one.
- Enter your own forecast for next cycle's total. The indices decide how that number is split across the year; they have nothing to say about how big it should be.
- Read the indices as multipliers of an ordinary season. 1.47 means that season runs 47% above average, 0.60 means 40% below. They are rescaled to sum to the number of seasons, which is what makes the season forecasts add back to your total.
- Check the disagreement between cycles before you trust any of it. If the two years put a season 0.2 of an index apart, something happened in one of them that has nothing to do with the calendar, and the average of the two describes neither.
- Take the peak season's forecast to your capacity and your safety stock — that is the number the warehouse and the cash have to survive, and an annual average never shows it.
What each term means
- Seasonal index
- What one season is worth against an ordinary one, as a multiplier. Above 1 is a strong season, below 1 a weak one, and they average to exactly 1 across the cycle.
- Cycle
- Whatever repeats — almost always a year. Two cycles is the fewest that can tell a season from a one-off, because one cycle has nothing to compare itself with.
- Deseasonalizing
- Dividing each period by its index, to see the series with the calendar taken out. It is how you tell a genuinely good December from an ordinary December.
- Multiplicative model
- Treating a season as a PERCENTAGE of average rather than a fixed number of units above it. A business that doubles keeps the same indices, which is why this form suits growing demand.
- Rescaling
- Nudging the indices so they sum to the number of seasons. Without it the season forecasts would not add up to the annual total they were split from.
- Peak to trough
- The strongest season's index divided by the weakest. It is the single number that says how much capacity, staffing and cash your calendar demands.
Educational disclaimer: Outputs are simplified educational estimates built from the numbers you enter — they are not financial, legal, tax, or investment advice, and real decisions deserve verified figures and qualified professionals.