Baresquare

The baseline is the reference point the actual performance of the revenue funnel is compared against. It can refer either to forecasts and targets set by the business (based on revenue plans provided) or to the expected behavior based on time-series analysis (in the absence of any revenue plans) and is the midpoint of the minimum and maximum expected values

The baseline used in each ticket is automatically selected by Tywin, depending on the availability of a Revenue Plan (and a respective expected revenue value for the reported product category), and defines the way that the min/max expected values are calculated: 

Revenue Plan (RP): When the Revenue Plan is available, and a target value for a specific product category can be retrieved for the examined day, the min/max limits are calculated based on this target value, as a ±10% difference from the target value (above and below).

Time-Series Analysis (TSA): If the organization’s Revenue Plan is not available or a target value for a product category cannot be retrieved for the examined day, the expected limits for revenue per product category are calculated through various techniques that examine the trend over the past 28 days before the examined day. The algorithm also checks the existence of seasonal trends (e.g. each Monday or during the weekend) that can impact the expected limits on those days. Essentially the min/max expected values are based on the normalized average figures during the previous 28 days after excluding outliers.

- Revenue Plan (RP): When the Revenue Plan is available, and a target value for a specific product category can be retrieved for the examined day, the min/max limits are calculated based on this target value, as a ±10% difference from the target value (above and below).
 
- Time-Series Analysis (TSA): If the organization’s Revenue Plan is not available or a target value for a product category cannot be retrieved for the examined day, the expected limits for revenue per product category are calculated through various techniques that examine the trend over the past 28 days before the examined day. The algorithm also checks the existence of seasonal trends (e.g. each Monday or during the weekend) that can impact the expected limits on those days. Essentially the min/max expected values are based on the normalized average figures during the previous 28 days after excluding outliers.