1. Abstract
Seasonal patterns such as holidays, weather cycles, or fiscal timing can obscure the true direction of a business metric. Applying the native seasonal adjustment functionality in Board Foresight removes these recurring effects, allowing analysts to focus on underlying trends and produce clearer, more reliable insights.
2. Context
Use this best practice whenever working with time-series data that exhibits predictable, recurring fluctuations. Seasonal adjustment should be applied before running Discover or using data in a Workbench to ensure consistent interpretation and correlation results.
3. Content
3.1 Why It Matters
Seasonality can easily be mistaken for growth or decline. For example, a December spike in retail sales may look like a structural improvement when it is simply a recurring holiday effect.
If seasonal patterns are not removed:
- Correlations with external drivers may be inflated or misleading
- Trend changes may be misinterpreted
- Forecasts may overreact to predictable calendar effects
Seasonal adjustment helps ensure that the signals you analyze reflect real changes in behavior, not timing artifacts.
3.2 How to Apply
There are two methods for applying a seasonal adjustment in Foresight.
Method 1:
- Open the indicator you want to adjust in Foresight.
- Launch the Transform Modal from the hamburger menu in the upper left.
- Select Seasonal Adjustment from the available transformations.
- Choose a Start Date and End Date for your seasonal adjustment period and choose whether to control for outliers.
- Select Update Preview to review your results or Apply to apply the changes to the selected data series.
- Use the adjusted version in Discover.
Method 2:
- Open the indicator you want to adjust in Foresight.
- Launch the Seasonality Test from the hamburger menu in the upper left.
- Choose a Start Date and End Date for your seasonal adjustment period and choose whether to control for outliers.
- If seasonality is found to be present, select Continue.
- Choose a Project in which to save your seasonally adjusted data series, and select Create.
- This method creates a new seasonally adjusted indicator and seasonal profile that you can use throughout your Board Foresight instance.
- Use the adjusted version in Discover.
3.3 Example
A retail analytics team applies seasonal adjustment to monthly sales data. Once adjusted, they discover that growth has flattened since Q2 — a trend previously hidden by recurring holiday spikes. This insight changes how leadership interprets recent performance.
3.4 Common Pitfalls
- Applying seasonal adjustment to some indicators but not others in the same analysis
- Confusing true structural changes (for example, post-pandemic demand shifts) with seasonality
- Double-adjusting data that is already seasonally adjusted at the source
- Failing to clearly label adjusted indicators
3.5 Expected Results
- Cleaner trend signals that align with business intuition
- More meaningful correlations in Discover
- Forecasts driven by true behavior rather than calendar effects
- Greater confidence when communicating insights to stakeholders
