A nice little case study on how important it is to track the RIGHT data for decision making

By: Hossein Jalali • August 20, 2015

A couple months back we had built a really neat little app for one of our clients and it was performing really well in terms of downloads.

In one of our meetings with the marketing team we were discussing the different campaigns they were running and how well each was performing. In those talks they mentioned that one of the main objectives of the app is to push new users to click on a promotional banner and fill up a form so the sales team can contact them.

They were complaining that Google was quite expensive and it cost them around $1.00 per app download while Facebook was giving them an average rate of $0.10 per app download; therefore they were deciding on shutting down the Google campaigns and putting that money on more Facebook advertising. I asked them to not take an immediate action and give me 2 weeks to run a few tests. 

After 2 weeks I prepared a presentation and met them again to provide some shocking results. We had placed a few more tracking codes on the form itself and the outcome was something like this:

Yes! Facebook brings in an app download for $0.10 vs the $1.00 of Google and if we leave it there the decision they made was correct; however out of 50 downloads from facebook only one actually fills out a form, on the other hand out of every 4 Google downloads one will fill the form.

The results portrays that actually the cost of filling a form (CPA) from facebook is $5.00 while Google is $4.00.

Therefore not only should they not stop the Google marketing budget on the contrary they should do exactly the opposite and try to move some of the budgets from Facebook to Google. 

This just proves how important it is to identify the right sources/KPI, a wrong choice can be quite devestating such as this scenario where the client was deciding to make a decision exactly the opposite of what was right just because they were looking at the wrong numbers. 

comments powered by Disqus