The 2015 break in conversion
Fixed.
The Bengaluru figures from 2010-2015 were sourced from a different agency than the 2015-onward figures, and while they overlap for one year (2015), the overlap shows a 12% discrepancy. This offset should be reconciled before using both sources in a continuous series. Should we either restate one to match the other's baseline, or should we split the series at 2015?
Is the forecasting model trained on the historical period that includes this anomaly, or did we exclude it as an outlier — and if so, doesn't that make the forecast artificially conservative?