Incidence of Reputational Periodicity

Reputation time series are analysed for periodic affects which might not be visible using simple smoothing techniques. To find possible frequencies that might indicate periodicity, a bandpass filter with a narrow passband, wide stopbands and steep roll-offs is progressively tracked on a day-by-day basis across reputation time series spanning a twoyear period. Dominant frequencies are recorded at each stage. The results indicate that periodic cycles exist in the general range of one to six months. This result is significant for reputation management because it indicates when upturns or downturns in reputation are most likely.