Here we use Python to analyze our Facebook Ad Results. We split our analysis into 3 buckets:
- Targeting (Age, Gender, Interests, Location etc.)
- Ad Operations (Device, Platform, Frequency, Day of Week, Time)
- Ad Appearance (Ad Format, Ad Copy)
You can set up these reports once and save them so that you can just change the timeframe and download these reports again without setting them up from scratch.
What insights we can gather:
- The effect of headlines, creatives and body copy on Age and Gender. You can use these insights to see what kind of Copy/Creatives are working on a particular Age/Gender.
- Similarly, you can see which combos of Ad copy, Ad Formats etc. work with a particular interest/audience segment.
- Next, we can see which day of the week/hour of launch is getting the best results along with Time-series data.
- Finally, we can check which platform is getting the best results (Facebook /Instagram /Messenger /Audience Network)
Naming Conventions:
For ease of Analysis, you can name your campaigns in the following way, with a hyphen (-) in between which can act as a separator while splitting up campaign names while analyzing.
Campaign Level:
Convention: Funnel Location-Objective-date
Eg: TOF - Conversions - 20/07/2021
Adset Level: