Data Analytics Case Study – Wrong Facebook Inflated Purchase Conversion Value Metrics 10X Times

I have seen a lot of Facebook marketers who promised to give you purchase ROAS up to 20X. Throughout my 7 years of online marketing career, I have hardly seen any display platform giving such a high return on ad spend. To the best of my knowledge, the highest Return on Ad Spend that I have seen is from branded search ads from the Google search network.

Are Facebook Metrics Inflated?

Yes, Facebook metrics are definitely inflated. I have analyzed around 5-6 facebook accounts for E-commerce businesses. All of these ad accounts were showing a minimum ROAS of at least 5X. The competitive campaign to such performance will be google branded search ads and none others.

So, I decided to dive deep into the data and tried to figure out the real ROAS, not the inflated ROAS. Below is the case study for one of the clients that I spent the most time on and used different platforms to see if I was doing the right calculation.

Case Study:  Facebook Inflated Wrong ROAS by 10x Times

So, the story began like this. The client was happy because Facebook was giving around 1.5X ROAS. The client hired me to scale up Facebook ads. As a part of the discovery routine, I decided to take a look at the overall performance of the paid channel in google analytics.

Note:

  • I have used the same lookback window of 30 days on all the platforms (28 days on Facebook as it does not offer 30 days)
  • I have used the last-click attribution model to attribute the conversion to right channels
  • In GA, all the channels have been defined properly.
  • In short, I like to have clean data.

Google Analytics Reported around 16000 in sales in total from all the paid channels in the month of September.

Revenue from Paid traffic

Facebook reported around 8,600 in sales from Facebook campaigns. 

Revenue from Facebook

Google Ads reported around 14,700 in sales from all of the Google campaigns. 

Google Ads overview

Inflated Numbers breakdown

The next step was to individually investigate each of the platform performance. So, I started with the analysis of the performance of Facebook Ads on Google Analytics

Google Ads reported around 200 in sales from paid Facebook ads

Facebook ads report in GA

If you see the picture above, as per Google Analytics, through paid Facebook ads, the total revenue was just around 200. WHICH WAS INSANE. So, I decided to use Facebook Attribution. A platform provided by Facebook itself. Here is what Facebook had to say for itself.

Paid Touchpoints on Facebook were credited for 20 Website Purchases in the last months from sources including Facebook and Instagram

Facebook Attribution Reporting

So now the picture was even clearer. For sure, Facebook was inflating its purchase value to the wrong number but, it was still unsure if the Facebook Attribution platform was closer or Google Analytics. So, I did another analysis for the month of August and September combined and here is what it looked like based on analysis from Google Analytics and Facebook Attribution combined.

Channels performance

  • Total Ad Spend: 20,458
  • Total Return: 1,897
  • Total ROAS: 0.09
  • Wrong ROAS reported on the FB Ads Platform: 1.53 (Almost 17x Higher). Screenshot Below:

Facebooks inflated ROAS

FIX:

  • Start using UTM parameters and use those parameters extensively to create custom channels.
  • Create Custom Channel Grouping in Google Analytics monitor the performance of different channels separately
  • DO not completely rely on FB ads manager data. Start using FB Attribution. It gives a lot of cleaner data compared to FB ads manager.
  • Besides Google Analytics and FB attribution, use another third-party tool such as Mixpanel to monitor the individual channel’s performance.
  • Compare total sales and revenue across different platforms regularly.

Leave a Reply

Your email address will not be published. Required fields are marked *