Trend

Forecaster calculates quarterly and yearly trend as a moving average performed over a period of time. Example trend intervals:

Quarterly trends are from the past 91 days
Yearly trends are from the last 365 days

Trend analysis requires populated values for at least 50% of historical data points.

Seasonality

Forecaster identifies seasonality. It checks for repeated patterns on weekly, monthly, or yearly timescales.

For example, to gauge a daily trend, Forecaster looks at the performance data for each Wednesday of the year.

Forecaster identifies seasonality patterns based on the historical data timescale. The table below describes the minimum amount of historical data you need for each type of seasonality analysis. Use its scroll bar to view seasonality options.

Even with positive historical values, you can get negative seasonality values. This is due to how the analysis was carried out:

The trendlines are subtracted from the historical data.
The seasonality data is calculated from the remaining historical data points, which can produce negative seasonality points.

If Forecaster doesn't identify any seasonal patterns, it can return a noisy output.

Tip: The more historical data you provide, the more accurate your seasonality analysis will be.

Results

To make the trend and seasonality results available:

Add these line items to the forecast results module:
- Forecaster_Seasonality_Weekly
- Forecaster_Seasonality_Bi_weekly
- Forecaster_Seasonality_Monthly
- Forecaster_Seasonality_Quarterly
- Forecaster_Seasonality_Yearly
- Forecaster_Trend_Quarterly
- Forecaster_Trend_Yearly
Follow the trend and seasonality instructions detailed in the Create an import action section of Stage 2: Create export and import actions. The results are accessible once the relevant forecast action runs successfully.

The trend and seasonality results are meant for analysis through visualization only, as the values aren't meaningful on their own. The results should be visualized in a chart along with historical data. We recommend that you use a secondary axis for ease of interpretation.

This video shows an example of how to create trend and seasonality charts.

title