How to leverage Google Analytics data to improve your Facebook campaigns
September 2, 2020
Using Google Analytics to improve Facebook paid traffic
- The problem with Facebook reports
- Segmentation: what’s this about?
The correct approach.
Questions to ask.
In the last few years, we’ve been constantly investing tremendous resources in building our own data analysis methodologies and best practices.
We have used processes and approaches that are practical and efficient when it comes to extracting actionable data insights. In our research, we tried to understand how to ask the right questions to find online behavior patterns, in order to use a template of questions that simplifies the process every time.
After working with multiple clients and seeing the same problems over and over again, we came to the conclusion that one of the most important questions you need to ask before analyzing any data is “Am I tracking the correct online data?”.
If you know your data is accurate, you can start leveraging it to get better results in your campaigns. If not, don’t worry, this guide will also give you some pointers on how to track it better.
Today we’ll focus on a specific area which can really impact your conversion rate: how to use Google Analytics to improve the performance of your Facebook Paid traffic. Specifically, we’ll show you how to do so by optimizing advanced segmentation.
Our idea for a custom template is very simple: get rid of all the code you need to create or delete a cookie and only fill a tag with all the information you want.
• Why Facebook Reports are not enough
• The correct approach to segmentation for best results
• Advanced segmentation: Finding Facebook optimisation opportunities in Google Analytics
What’s the issue?
Compared to other marketing channels, Facebook Paid campaigns tend to have a higher bounce rate when you don’t put in the effort to optimize both the campaigns and the website. Your traffic might grow, but the leads you get are not qualified so most of your money is being blown out the window.
You might already be testing and optimizing your Facebook campaigns, but if you’re relying only on Facebook’s performance reports, you are missing out on what is actually happening on your website.
The problem with Facebook reports
Facebook does a decent job on informing you about the performance of your campaigns, it even gives you a few details about your target audience, however, it does not provide you with in-depth information about the behaviour of people who land on your website. This information is crucial to understanding what you can optimise and how to get more bang for your buck.
But now that you know the full picture, you also know that Facebook would have never made it alone. So if you only use your Facebook Reports and do not keep track of what happens on your website via Analytics, you’re missing the winning strategy for that sale and how to scale it.
That’s what we’re going to explore today.
Segmentation: what’s this about?
Above is a sneak peek of how most Advertisers use Google Analytics.
Don’t get me wrong, this can be helpful too, especially if you’re not only looking at all website data lumped up together but are instead looking at Facebook campaigns only or Facebook Campaigns compared to other sources.
But that’s just the tip of the iceberg, and as with most icebergs, if you don’t pay closer attention, it can sink your ship. So let’s look a bit deeper and understand how to avoid sinking.
When thinking of website traffic, we can see in Analytics it comes from multiple channels, one of which is Paid Traffic.
Paid further splits into paid search, paid social and display. Now, they might look alike, but they are very different advertising channels and your ad efforts should take this into account.
Let’s take the Paid search.
In a nutshell, Paid Search traffic comes from people who searched for something in Google, saw one of your ads and clicked on it. This person knows what they want, and they intentionally searched for it, as opposed to somebody who just saw a promoted ad without intentionally asking for it.
Someone who is searching for something specific on a search engine is in an advanced stage of awareness. This person knows their pain and is actively trying to solve it, meaning she is closer to converting, as opposed to those who see promoted ads on Facebook, Twitter and the like.
So this segment of people acts differently on your website, as opposed to those who come from Facebook (Paid Social). Click the infographic on the right to zoom the differences.
Let’s start with showing you how you can separate this channel in Analytics and properly see the results you get from it.
The correct approach to segmentation
You might have heard that it’s mandatory to add UTMs to your Facebook campaigns in order to properly track which of your links is bringing what leads, and you might have done it in different ways.
We strongly recommend creating a custom channel for your paid social traffic and analyzing it separately from your paid search. We already mentioned above how people who come from Paid Search (google ads) are in a different state of awareness than those who come from Paid Social (facebook ads).
So, as mentioned, a good UTM structure will allow you to pinpoint which one of your campaigns is bringing in the best leads, making it easier to understand what is happening in Google Analytics.
The best UTM structure is:
Let’s have a look at how Google defines the Paid Search channel:
This means that if you tag your paid social traffic with medium PPC (Facebook, Linkedin, Instagram, etc), it will, for example, get lumped up in the same channel as your paid search traffic from Bing or Google Adwords. Choosing this option might render your optimizing efforts null.
Instead, if you are using the UTM medium paidsocial you can go ahead and build a custom segment.
TIP: If you have already created a custom channel for Facebook paid traffic, well done! If NOT keep in mind these 2 things:
• To get data from the past create a custom segment
• To have your future data better organized, and create reports much easier, create a custom channel
If you want to understand your customer’s mind and spot behavioral patterns on your website, segments are your magic tool.
It’s all about looking at a segment and then at a smaller subset and then at even more granular segments until it gets actionable and meaningful. It might sound tedious, but when you do it right, with a few clicks you get the valuable data that power your growth.
Going back to our previous idea, that is why you first have to isolate your paid social traffic, apply this segment to your reports and then dig deeper and play with additional custom segments.
Finding Facebook optimisation opportunities in Google Analytics
Now that we know the basic rules on isolating only the traffic from Facebook campaigns, we can go ahead and look at more advanced segments.
We’ll do that by asking specific questions.
Q #1: How are the top Facebook landing pages performing?
Let’s imagine we’re looking at the clothes e-commerce website from the example given earlier. One of the first questions that you can ask is:
Where do you send the traffic from Facebook campaigns and how are these pages performing?
I want to emphasize here, go granular:
• First, most of the times, it makes sense to separately analyze data by device, by gender, by the number of pages viewed and so on in order to spot behavioral patterns.
• Second, don’t only look at macro goals (conversions), but also look at micro goals (newsletter subscription). People don’t buy from you straight away, but there are some solid indications which predict a higher chance that they’ll do that.
Ok, so let’s see how this data is going to help us.
Q #2: What pages macro-convert the best/ worst?
This is the exact table as the one before, but we start asking more specific questions to spot online behavioral patterns.
We can see a couple of things:
• Category Pages have the highest conversion rate, both on mobile and desktop.
• The deals page bounce rate is significantly above average on both devices (75%), having the worst conversion rate.
• The category page for female performs the best in both conversion rate and average order value ($164)
• The conversion rate of category pages is 4 times higher for returning visitors compared to new visitors.
Which leads us to the following:
By setting up category pages as landing pages for your re-marketing campaigns, we expect a higher return on investment.
You start with these types of insights, but you want to get even more specific with them. So you continue the segmentation process by asking questions such as: “Is there a significant difference between new and returning visitors who land on category pages and are coming from Paid Facebook campaigns?”
This is how you are coming up with hypotheses that have a high winning chance when AB testing them.
For example, here we already know that the deals page’ bounce rate is 75.99%.
But after taking a closer look at data, we also find out that certain age groups perform significantly better than others. So the hypothesis is focused on targeting these groups for a higher ROI.
• The deals page bounce rate is significantly above average on both devices (75%), having the worst conversion rate.
• But when we dug in deeper to look at how different age groups convert on deal pages, we observe that two age segments [18 -24] and [25 -34] have a pretty good conversion rate (1.5% and 2%)
Which again, leads to the following:
By targeting only visitors of age between [18 and 34] and sending them on deal pages, we’d expect a higher return on investment.
Q #3: What pages micro-convert the best/worst?
But as I was saying earlier, only looking at macro-goals is not enough. The amount of micro-goals we are looking depends on the website complexity and the amount of traffic, but here I chose the following two: product search rate (website search bar) and newsletter subscription rate, which might be more relevant especially for new visitors.
• The search bar rate is significantly higher for category pages.
• The newsletter subscription rate is significantly higher on homepage, both on desktop and on mobile.
• The email conversion rate is 6.5%
By sending new visitors to your homepage, where your main call to action would be related to “Newsletter subscription” (first-time purchase offer), we expect an increase in return on investment.
So we know the majority of the newsletter subscriptions come from homepage, but are those emails efficient?
Well, in short, yes.
The email conversion rate is 6.5%.
Q #4: What are the main differences between mobile and desktop?
We are keeping an eye out for specific behavioral patterns by device:
• The average order value is 50% higher on desktop compared to mobile.
• The average number of products purchased on mobile is 1.1 whereas on desktop is 1.75
• The age groups that convert better on mobile are [18-24] and [25-34]
• The bounce rate on mobile is very high, 75%
#1: Given that mobile visitors tend to buy 1 product per purchase, we expect an increase in conversion rate if we’d give up the cross-selling section on the checkout pages.
#2: We expect a better return on investment rate by targeting on mobile only people of age between [18 and 34].
To wrap up, never lose the big picture out of sight. If fully understanding the customer journey and how to influence it is what you’re after, then count on something more than just Facebook’s reports.
It’s been said over and over, by many analytics evangelists and not only, that segmentation is what gets you actionable insights. And that’s not just a buzz phrase, it’s the reality that brings you results.