Why does a Gnip PowerTrack Twitter estimate with a shorter time duration have more tweets and cost more than for a longer duration?


I submitted two Gnip PowerTrack Twitter estimates that confuse me. I have shorter time window on one, but the shorter window led to a higher count and cost. The search criteria is identical for both estimates. For example:

Estimate #1 - shorter duration, more tweets

Start Date: 08/11/2014
End Date: 08/12/2014
Estimated Activities: 407,000

Estimate #2 - longer duration, fewer tweets

Start Date: 08/11/2014
End Date: 02/11/2015
Estimated Activities: 173,000

Could you explain why this is?


This is a by-product of the particular methods used by Twitter for generating estimates. If the main activities for a day fall outside of the estimate window, then they will not be fully reflected in the estimate.

We advise taking a variety of estimates with different date range combinations as a work around. This usually solves the problem.


Have more questions? Submit a request


Powered by Zendesk