Adventures with R - Facebook Ads (Part 1)
A new year resolution in 2018 has been trying to learn new data science technologies.
An attempt over the previous few weeks has been trying to master R. R is an open source programming language supported by the R foundation for statistical computing. Originally used by statisticians, it has now found applications in nearly every field of machine learning.
One of the interesting applications at work was trying to automate Facebook Ads. Facebook has an advertising platform that uses advanced optimization techniques to serve the ads to the right audience at the right time. The challenge has been getting a daily feed from Facebook about who sees the ad, who interacts with it and who ends up buying your end product on an ecommerce site.
There have been three main ways I tried extracting data from Facebook
1. Facebook Business Manager Login : Manually logging in and getting the data out. This was a cumbersome approach and not prone to easy automation.
2. Blendo.co : Blendo is a world class integrator. You add your facebook business manager credentials, your cloud database credentials, set up a sync frequency schedule and you are good to go. The issue I ran into is budget - I dont really have a big budget and Blendo has a pricing structure (https://www.blendo.co/pricing/) that is outside my budget limits.
3. Analytics Edge Excel Add-in : This has been my go to product for several months now. A very good tool that is extremely easy to set up and use. Outside the trial, it will cost $50 per year but it has excellent customer service (emails answered within a day or less) and data that matches Facebook 100% of the time. https://www.analyticsedge.com/simply-free/
I did want to use R as a technology demonstrator. We will continue using the Analytics Edge add in as it has wider applications outside analysts but this series of blogposts is around using R to automate Facebook ads.
Stay tuned for part 2 !