One month is too narrow to prove seasonality, especially if new-listing volume changed during the same period. I’d split the sample by neighbourhood and condition, then track new, reduced, withdrawn and completed properties separately. If reliable completed prices are unavailable, treat 96 days...