Seasonality indices from history
Counts 12 monthly seasonality indices from sales over 2–3 years: each month is divided by the monthly average of its year, the indices are averaged and brought to a sum of 12. Shows the span of seasonality and spreads the annual plan across months.
At a glance
- What question does it answer?
- What seasonal shape the sales have by your history and what share of the year falls on each month.
- What data is needed?
- Sales by month over 2–3 years of one measure.
- How to read the result?
- A span below 15% means there is no seasonality. A large divergence of the year profiles means the shape is unstable. Example on the default values: Month with the maximum index: December; Maximum index: 1.88; Month with the minimum index: February; Minimum index: 0.65. The figures are illustrative: enter your own to see your result.
Why and how to reason
The annual plan by month is spread by the shape of the year, and the shape comes from history. Each month is divided by the monthly average of its year, the indices are averaged and brought to a sum of 12. In the example there are two years: the maximum is in December 1.88, the minimum in February 0.65, the span is 123% of the average.
When to use it
Apply when drawing up the plan for the year, when spreading the budget by month and when cleaning the rate of sale with a seasonal adjustment.
How to read the result
A span below 15% means there is no seasonality. A large divergence of the year profiles means the shape is unstable or there was an anomalous year.
Common mistakes
One year is taken, a year with an anomaly is not excluded, revenue and units are mixed when prices rise.
What data is needed?
- Sales by month over 2–3 years of one measure.
- A mark of anomalous years.
- The annual plan (optional).
Inputs
| Field | Unit | Value in the example | Meaning |
|---|---|---|---|
| Year 1 | Use | Exclude a year with an anomaly (a one-off promotion, shortage, closure): it distorts the indices. An excluded year does not take part in the calculation. | |
| Year 2 | Use | Exclude a year with an anomaly (a one-off promotion, shortage, closure): it distorts the indices. An excluded year does not take part in the calculation. | |
| Year 3 | Exclude | Exclude a year with an anomaly (a one-off promotion, shortage, closure): it distorts the indices. An excluded year does not take part in the calculation. | |
| Annual sales plan for next year (optional) | in the data units | 120,000 | In the same units as the sales in the table (roubles or units). If given, the plan by month is shown. Empty: the plan is not spread. The example value is illustrative. |
How it is calculated
| Step | Formula |
|---|---|
| Ratio to the average | r(year, month) = sales of the month ÷ the monthly average of that year |
| Index of the month | the average of r over the years used |
| Normalization | the indices are brought to a sum of 12 (to 52 for weeks) |
| Span | (maximum − minimum) × 100% of the average index 1 |
| Plan of the month | annual plan × index ÷ 12 |
Worked numbers
Default values from the app; in the interactive version you enter your own.
- Month with the maximum index
- December
- Maximum index
- 1.88
- Month with the minimum index
- February
- Minimum index
- 0.65
- Seasonality span, % of the average
- 123 %
- Sum of the indices
- 12Always equals 12.
- Years used
- 2
- Divergence of the first two year profiles, index per month
- 0.02
Data as a table
| Measure | Index |
|---|---|
| Jan | 0.74 |
| Feb | 0.65 |
| Mar | 0.81 |
| Apr | 0.93 |
| May | 0.87 |
| Jun | 0.83 |
| Jul | 0.79 |
| Aug | 0.83 |
| Sep | 1.01 |
| Oct | 1.19 |
| Nov | 1.48 |
| Dec | 1.88 |
Indices and the plan by month
| Month | Ratio to the average of year 1 | Ratio to the average of year 2 | Index | Share of the year | Plan of the month |
|---|---|---|---|---|---|
| January | 0.76 | 0.72 | 0.74 | 6.1 % | 7,364 |
| February | 0.64 | 0.65 | 0.65 | 5.4 % | 6,471 |
| March | 0.81 | 0.81 | 0.81 | 6.7 % | 8,095 |
| April | 0.93 | 0.93 | 0.93 | 7.7 % | 9,273 |
| May | 0.87 | 0.87 | 0.87 | 7.3 % | 8,704 |
| June | 0.85 | 0.82 | 0.83 | 6.9 % | 8,325 |
| July | 0.77 | 0.81 | 0.79 | 6.6 % | 7,906 |
| August | 0.83 | 0.83 | 0.83 | 6.9 % | 8,312 |
| September | 0.99 | 1.02 | 1.01 | 8.4 % | 10,052 |
| October | 1.19 | 1.19 | 1.19 | 9.9 % | 11,899 |
| November | 1.47 | 1.48 | 1.48 | 12.3 % | 14,783 |
| December | 1.89 | 1.88 | 1.88 | 15.7 % | 18,816 |
Two years are used: the indices are steadier with three, if there is no anomalous year.
The 11.11 and Black Friday events, 8 March and school holidays are inside the monthly index; if an event moved between months the index of the month is distorted. The weekly profile is counted the same way with normalization to 52.
What to do with the result: spread the annual plan by month and the receipts tape (wosreceipts) by the indices; an index does not replace the growth or the fall of a year, it describes only the shape within a year.
A typical mistake: taking one year, not excluding a year with an anomaly and mixing revenue with units when prices rise.
Nuances and limits
- Taking one year.
- Not excluding a year with an anomaly.
- Mixing revenue and units when prices rise.