The power of Google Analytics in combination with Facebook
Recently I conducted some in-depth attribution analysis for one of my clients.
The analysis was around proving the value for Facebook. The platform was used for implementing a marketing funnel and I tried to determine conclusively if this Facebook campaign works for this client. When I analyzed the data via Google Analytics in more detail, I couldn’t see that much of an overall impact on the business, so naturally I wanted to dive a little deeper.
Most people (including me) need to see improvement in a pre-defined metric to believe if something is working or not. If you are spending $20K a month on a given channel you want to make sure your money is well spent.
So, when you analyse Facebook performance there are probably a hundred metrics you can look at. So, which ones do you chose?
First of all, it largely depends on the purpose of the campaigns. The metrics will be different for a pure brand awareness campaign versus a remarketing campaign that advertises to previous customers / website visitors.
In my case, I analyzed upper to lower “funnel” activity of the user journey (acquisition and remarketing) but the main focus was on acquisition as I wanted to understand if this channel actually works from this point of view.
When I analyse data I usually measure performance by utilizing real data + common sense. You can interpret data in different ways, therefore, making the obligatory common sense check is vital.
Google Analytics (GA) is a great tool to measure performance. If you are just getting started with GA, here’s a beginner guide.
- In this particular case, GA would have likely missed some touchpoints – touch points I needed to understand to know if the platform is good, bad or ugly… so I went onto Facebook to analyze the performance in more detail.
If you have certain expectations and targets to achieve but GA just shows a handful of conversions from your Facebook acquisition campaigns, you might want to investigate further. Why? Because Google Analytics is excellent in most things regarding data, but sometimes a symbiosis between GA + Facebook (or other platforms) gives an even better picture.
So, I went to Facebook and had a closer look at view-through and click conversions. Here’s a really comprehensive guide on this topic.
- As a default you will find a 28-day click and a 1-day view conversion window for Facebook.
- The default is set to 28 days. But you can play around and see what works best for you. In my case, I wanted to prove the value of Facebook by looking at 1-day views and 1-day click conversions. Please note that Facebook kinda hides view-through and click- conversions in one column so you will need to tick the boxes
… and add different conversion windows to compare.
- Whereas clicks from Facebook will be recorded in GA by looking at the corresponding source / medium, Google will not give you any data on Facebook view-through conversions.
- We are aware of the fact that view-through conversions are a highly controversial topic. Some marketers disregard this metric entirely. Other marketers give great value to view-through conversions.
Personally, I feel, view-through conversions should not be disregarded entirely. I am rather conservative when looking at this metric but I still think that there is some value in it.
Please see below how Facebook defines a view-through conversion
Please also note the below:
Active Facebook users
If you think about 1-day view conversions a bit further, we know that
- Roughly 1.66 billion people on average log onto Facebook daily and are considered daily active users for December 2019. (Source)
- And Worldwide, there are over 2.50 billion monthly active users for December 2019 (Source)
Why is this important?
Because a very large amount of people actually use Facebook on a regular basis – meaning, the likelihood of anybody seeing your ad is pretty high. Therefore, you want to shorten the conversion window as much as possible as you want to measure the direct impact of an ad to a sale.
You probably want the view-through conversion window to be the minimum of 1 day if possible. Longer conversion windows (e.g. 7 or 28 days) are more likely to “dilute” the direct impact of your ad on a sale.
1-day conversion window
When I had a look at the 1-day view conversion value for this specific client the numbers were huge. Now, if you can see $1 million in revenue attributed to your campaigns by looking at a 1-day view conversion window, how do you feel about this AND how much do you think Facebook contributed to this revenue in reality? Did people even consciously look at the ad?
Well, this number is probably the trickiest part, as it is difficult to say for sure. I have seen people estimating 10% up to 40% based on calculated guesses (whatever reasons they had).
Let’s say in our particular case we estimated 15%, which would equate to $150,000 in assisted conversion value from mainly upper-funnel activity for a cost of $20K. Would you turn this ad off? It will surely depend on your goals and KPI’s of course.
In my case, we wanted to figure out if Facebook was delivering superior results to other channels and by looking at the data, we could prove the value of Facebook.
But you may want to have a look at other metrics and use Facebook to achieve other goals:
- Limitations reached in existing channels
- Maximising targeted reach
- part of a multi-channel approach
Although Google Analytics is an awesome tool for data analysis, it doesn’t hurt to look at attribution from another angle to improve the quality of your insights. If you silo yourself into just one view, you could miss the bigger picture.
In the end, each platform has its strengths and weaknesses and we all want to maximise growth for our clients cost-effectively.