Session Ready
Exercise

Setting up conversion indexes

Now that you've determined that language is, in fact, the issue with House Ads conversion, stakeholders need to know how many subscribers they lost as a result of this bug.

In this exercise, you will index non-English language conversion rates against English conversion rates in the time period before the language bug arose.

Instructions
100 XP
  • Create a new DataFrame, house_ads_bug, that contains only the rows from house_ads with 'date_served' prior to '2018-01-11'.
  • Use your conversion_rate() function on the house_ads_bug DataFrame and the 'language_displayed' column.
  • Using the appropriate row, divide the Spanish, Arabic and German columns of lang_conv by the English column.