1
$\begingroup$

Background:

I have a dataset with an event that started on November 28th, 2017 at specific locations (target). The goal is to measure the success of the event by comparing the same metric against the year-over-year trend.

My initial thought was to look at the sessions for the same date range year over year: Nov 14, 2016 - Dec 10, 2016 vs Nov 14, 2017 - Dec 10, 2017 using the November 28th as a point of reference since this is when the event started.

The problem:

Cyber Monday has an influx of sessions both last year and this year. Last year, Cyber Monday fell on the 28th and this year it fell on the 27th; therefore, last year's volume looks much higher when comparing the daily figures to this year which doesn't include Cyber Monday.

Further thoughts/questions

  1. Should I remove Cyber Monday data from both this year and last year?
  2. Do I need to compare year over year trends to find the impact of the event?
  3. Outside of the context of this problem, what is the best practice when comparing two data sets when one has a known outlier (Cyber Monday)?

Data set:

Event,Date,Non-Target Sessions,Target Sessions
FALSE,14-Nov-16,847815,51258
FALSE,15-Nov-16,959918,50370
FALSE,16-Nov-16,714895,52098
FALSE,17-Nov-16,645442,50875
FALSE,18-Nov-16,540389,45172
FALSE,19-Nov-16,573212,34365
FALSE,20-Nov-16,526001,34484
FALSE,21-Nov-16,778588,68712
FALSE,22-Nov-16,755597,68198
FALSE,23-Nov-16,762425,67974
FALSE,24-Nov-16,699033,59484
FALSE,25-Nov-16,824158,99285
FALSE,26-Nov-16,541329,57489
FALSE,27-Nov-16,527835,50007
FALSE,28-Nov-16,860053,101462
FALSE,29-Nov-16,625637,63732
FALSE,30-Nov-16,643841,63520
FALSE,1-Dec-16,571777,55826
FALSE,2-Dec-16,517024,49443
FALSE,3-Dec-16,390715,33810
FALSE,4-Dec-16,405985,34150
FALSE,5-Dec-16,566815,51204
FALSE,6-Dec-16,567957,50536
FALSE,7-Dec-16,619877,54373
FALSE,8-Dec-16,617475,51851
FALSE,9-Dec-16,506037,42769
FALSE,10-Dec-16,391132,33306
FALSE,14-Nov-17,430069,46172
FALSE,15-Nov-17,419564,44246
FALSE,16-Nov-17,416758,45140
FALSE,17-Nov-17,372086,39803
FALSE,18-Nov-17,294428,31685
FALSE,19-Nov-17,301626,30942
FALSE,20-Nov-17,471373,54158
FALSE,21-Nov-17,490153,55499
FALSE,22-Nov-17,521766,60652
FALSE,23-Nov-17,508995,61455
FALSE,24-Nov-17,564945,71765
FALSE,25-Nov-17,379483,47683
FALSE,26-Nov-17,385061,46093
FALSE,27-Nov-17,713916,92884
TRUE,28-Nov-17,480300,56529
TRUE,29-Nov-17,465913,52995
TRUE,30-Nov-17,436258,49997
TRUE,1-Dec-17,415525,50141
TRUE,2-Dec-17,293033,33392
TRUE,3-Dec-17,311525,33958
TRUE,4-Dec-17,443021,49821
TRUE,5-Dec-17,440373,49286
TRUE,6-Dec-17,442250,50063
TRUE,7-Dec-17,422875,48637
TRUE,8-Dec-17,392362,45155
TRUE,9-Dec-17,262810,61805
TRUE,10-Dec-17,310304,32816
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.