Can you imagine being able to track the ROI of traffic from…
… and more? You need Google Analytics campaign tracking.
When you have the ability to track your ROI on every launch, every funnel, every email and ad campaign — you can cut the fat and double down on what’s working.
This transforms your business from one that spends time, money, and resources on strategies that just seem like they’ll work well, to a business that makes smart, data-driven decisions and knows what strategies will work well..
Today we’ll get you started with campaign tracking with the first two (super critical) steps…
Ready to get started? Let’s start with tracking how visitors are getting to your site. We’re going to do this with UTM parameters…
Do you know what UTM stands for?
If so, you’re officially one of the cool kids in digital marketing. UTM is short for…
Urchin Tracking Module, a system that allows users to tag hyperlinks in order to trace where visitors originated.
If you’re a Google Analytics user (which you should be), you can use these to figure out how people are getting to your site (and what they’re doing when they get there).
By adding additional text on the end of every hyperlink you share, you can tag people who click those links (and hit your site) with relevant information about…
Let’s give a quick overview of the different UTM parameters that matter:
Generally speaking, source describes where you visitors came from.
It tells you the specific place where the referring link lives:
Common sources include:
Medium tells you how the visitors got to your site.
This is the highest-level way to sort data with utms, and consequently includes the broadest categories. Some of the most common mediums include:
This describes the specific ad, banner or email used to share the link.
It is used to determine what creative is working best at promoting an offer or distributing content. This will vary wildly by platform, content type, and offer, and there’s no standard for nomenclature.
My advice? Be descriptive so that you can easily remember what email or ad you’re talking about.
Campaign is similar to content, in that it’s a pretty open-ended field.
Its basic purpose is to highlight promotional offers or content distribution strategies so that you can easily compare performance across time and platform.
Campaign links should be consistent across all different sources and media for any given promotion to ensure the campaign as a whole can easily be analyzed.
Here’s an example that shows exactly what one of these links might look like for a flash sale of the Content Engine (one of DigitalMarketer’s Execution Plans):
(RELATED: Want More Social Media Traffic? Follow this 6-Step Blog Content Distribution Plan)
Creating properly attributed hyperlinks will take some time to get used to, but the data it provides will be worth its weight in gold. To make consistency easy, I recommend creating a unified document where you track all the hyperlinks you use, making it easy to refer back to when you’re analyzing later.
This is what mine would look like….
Luckily for you, Google makes actually building these links super easy! They have a free UTM builder where you can just plug in your information and automatically generate a hyperlink.
Now we’re going to see what people do when they arrive on your site. Slightly more complex, A LOT more exciting.
Goals are a way to track the actions people take on your site by tallying specific behaviors.
What makes goals really useful is not just the ability to say how many times an action was taken, but to look at who took the action. Thanks to the UTM parameters, you can actually do this!
But let’s not put the cart before the horse — let’s go over how to set up the most basic goal: opting in for a lead magnet.
Let’s build one to track our Perfect Blog Post Template.
We want to know when someone hits our Thank You page, after they’ve visited the opt-in page. Here’s how to make this happen:
Google offers a variety of goal templates, which should fit your specific needs (though you can create custom ones as well). Since we want to track opt ins, ‘Sign Up‘ should be perfect for us.
For Destination, change your rule to ‘Begins with’ and add your Thank You page — that’s where people who opt in end up.
Using ‘Begins with’ helps ensure all opt ins are properly credited.
The other way to ensure that we’re tracking actual opt ins and not just accidental Thank You page visitors is to create a funnel. This involves adding the URL string of the opt-in page as well (check out the graphic below). You’ll turn Funnel to “ON” and add a step with the page field including the URL that precedes your destination page. Set this step to Required and you’ve added this rule!
For your URL strings, always use the text after your domain name, not the entire URL — Google already knows the root domain!
Once you’re done, verify your goal to make sure you set it up correctly. Then click Save and you’ve built your first goal in Google Analytics!
Since we just set this goal up, we’re not going to have any information to look at. So let’s review an older funnel already tracked with goals and see how we can use these goals with UTM parameters to get insight into our customers.
To take a peek at these insights, you’ll want to visit Conversions > Goals > Overview under the reporting section of Google Analytics. We select a goal from the dropdown at the top and then near the bottom we’ll change the details from Goal Completion Location to Source / Medium.
Here’s what we see:
You can see that Facebook ads are one of our biggest sources of traffic. Interestingly, it looks like our automated followup emails are the next biggest source of conversions. Goals are a great way to get insights into what channels are driving the most visits.
With a little bit of know-how, and a lot of proper attribution tagging, Google Analytics campaign tracking can give you some great insights!
Now that we know the bare bones needed to properly track success and what channels are driving that success — how do we put this to work?
Like I said earlier…
When you have the ability to track your ROI on every launch, every funnel, every email, and ad campaign — you can cut the fat and double down on what’s working.
In Part 2, I’m going to show you exactly how DigitalMarketer cuts the fat and doubles down on what’s working through unique strategies combining proper attribution with audience insights.
This powerful information gives us the ability to break our audience up into segments based on…
In the context of analytics:
Segments represent groups of visitors grouped by shared characteristics or behaviors.
For the example we’re looking at today, conversions will represent Lead Magnet downloads.
Let’s take a look at…
The paid traffic campaign drives traffic to this landing page…
The campaign has produced 14,764 leads so far, giving us plenty of data to get a sense of exactly who is interested in this offer and what channels are working best.
If 15k leads sounds impossible, don’t worry — you won’t need this many leads to start getting actionable data. We are going to cover how to create useful segments, and how to drill down into these segments to better understand the people taking the actions we’re looking for.
There are 2 main goals in this analysis:
Sounds pretty simple, but this is one of the best ways to make the most of your budget, or know where you should spend your time and energy driving new customers.
Let me introduce you to your new best friend:
You’ve probably seen this when you’ve logged into Google Analytics.
….But I bet you haven’t spent too much time looking at this information.
And even if you have, I can promise you that it hasn’t been actionable.
That’s because the default is to aggregate all site visitors into a single audience, which leaves you with a mush of different groups all mixed together into an unusable mess.
Our strategy will show you how to focus on the most valuable, highest-converting audiences, and figure out what makes them tick.
Before we dive into exactly what information you should be looking at, and how to pull it, let’s talk about how to create the segments you’ll use to dig into your audience information.
There are 2 key strategies you can use to build these, the first of which relies on the campaign tracking we presented in Part 1. Luckily, if you’re still feeling a little confused about campaign tracking, we’ve got a workaround you can use to get some of the same information — as long as your attribution is correctly done.
Creating segments is fast and easy — you create a set of rules that include or exclude certain people, allowing you to narrow down your audience to look at a specific subset, like people who opted in for a Lead Magnet, rather than all site visitors.
Once you’ve got your segment created, you can analyze how this subset of visitors behaved or, in our case, who is in the subset. Let’s walk through exactly how you can create your own segments.
First, click “+ Add Segment” which you can find under the Audience, Acquisition, Behavior, and Conversion sections.
Once you open up the segment menu, we’re going to create new segment.
We have 2 options for how to create our segment…
The most accurate is to combine Goal Completions with traffic source, which gives you a breakdown of site visitors who took the action you want to analyze — in this case, opting in for the 10-Minute Social Media Audit — that came from a specific channel.
If you haven’t been using goals, you can alternatively look at site visitors who reached the Thank You page for the opt-in.
This is typically less accurate and does not work if you direct traffic to the Thank You page from any other source, so the best choice is to use goals.
Let’s walk through how to actually set this up.
First, we’re going to go to Conditions. This is where we’ll select goal completions, or page visits, as one of our audience creation rules.
If you’re using Goals as your inclusion rule, create a condition that includes Users who have completed the Goal more than 0 times.
…Yes, Google Analytics terminology sure sounds weird sometimes.
If instead you’re using page visitors, you’ll set users to include page that contains the slug of the URL.
Now that you’ve got your condition set, you’ll go to Traffic Sources and include the channel you want to focus on.
For us it’s going to be Facebook as the source and PPC as the medium, so we can specifically see how our Facebook PPC ads performed.
Now that we’ve created the segment, we can see just how big it is:
So we have 6,202 users in this segment — that’s 6,202 people who opted in for our 10-Minute Social Media Audit they found through a Facebook PPC ad.
That’s more than enough users to make this work.
As a general rule, I think you want to aim for a minimum of 3,000 people in your audience — that ensures you’ll have reasonable enough subgroups to put some faith in your age demographic groupings.
You can experiment with fewer, but the more conversions you’ve got in a category, the more trustworthy your data will be.
Now that you’ve got your segmented audiences, let’s dive into the meat of this post…
How to figure out who exactly is opting in (and who’s not), so you can do more of what’s working.
Understanding your audience means driving down ad costs, or figuring our better strategies for monetizing the leads you’re getting! So, if you’re interested in spending less OR making more, you should probably read on.
There are 2 main types of data we’re going to be looking for:
Demographic data describes what people are like.
Generally it’s statistical details like age and gender, but we’ll be lumping device type and location into this category as well.
This information helps you understand exactly who you are speaking to, and will guide the targeting you use in your campaigns.
Psychographic data instead describes what people like.
It’s all about interests, hobbies, and likes and can speak more to the personality of the audience. Psychographic data is most powerful when it’s used to shape your messaging.
We’ll be looking at a few different ways Google Analytics defines interests, including affinity categories and in-market segments. We also are going to include an estimate of the audience’s general wealth.
Normally household income is demographic, but because we’re only going to be able to say whether or not the audience is in the market for luxury goods, it’s less hard data and more of a way to feel out what they are looking for.
Here’s a quick breakdown of the data types:
Let’s start with the demographic data — it’s pretty cut and dry (though yields very interesting insights).
We’re going to actually create custom reports, rather than use the actual insights tab. That’s because we need to make sure we look at Users, not sessions.
However, we’ll be tracking all the data points you normally look at in the audience insights tab.
For our demographic reports, we want to look at age, gender, location, and device type.
To get started, go to Customization and select “+ New Custom Report.”
We’re going to build one report to look at demographic data, and one to look at psychographic data. That will make it easy to run these reports for all kinds of audience segments.
Create 4 Report tabs, one for Age, Gender, Location, and Device Type.
You’ll set the Metric to Users for all 4 of these reports — that’s the constant between them.
For Age, the dimension drill down will be Age, for Gender, it will be Gender, and for Device, it will be the Device Category.
Location is the only oddball; instead of doing a normal explorer view, you’ll use map overlay (which replaces the need for a dimension).
Save the report when you’re done and you’ve got a ready-made demographics report to use time and time again!
Still with me? Now let’s look at the data.
We’ve still got the segment we created applied, so we can dig in and see exactly who is responding to our 10-Minute Social Audit.
Here’s a quick overview of the age and gender:
At first glance, we can see that this campaign is working well with both the 25-34 range, then the 35-44 range.
It’s particularly interesting because a lot of these ads were aimed at supervisors, which would most likely skew the age average upwards.
But the fact that the 25-34 age range dominated with over 30% of the opt-ins suggests that either the ad appealed more to non-bosses OR the average age of bosses that care trends younger.
Either way, it’s likely that the ad copy needs to be re-evaluated through the lens of a younger audience, because while 35-44 was the second most prevalent age group, it ate up a larger percent of the overall campaign budget.
Retooling the campaign so that it speaks to a younger audience could do a lot to increase opt-ins and drop cost per click (CPC).
As for gender, women represented nearly 2/3 of the opt-ins for this campaign, which is a significant enough percent that the ads should be optimized for a female audience.
Since these ads weren’t weighted to allocate more budget to either gender initially, creating a second version of the campaign that solely targets women is a great way to boost clicks and ad relevance, and drop CPC.
Location isn’t particularly interesting for this campaign, because we only targeted the US, UK, Canada, and Australia with our Facebook ads. If you were looking at a different channel, this would definitely be a data set worth exploring.
You may realize that spending your traffic budget in a country you didn’t consider could really take advantage of regional interests.
One of the most important things for DigitalMarketer to note is the Device split:
This is so critical because it’s radically different than what we see across the entirety of our site, and audience segments generally speaking.
On average, about 33% of traffic that comes to our sites is from mobile.
However, for the people interested in the Lead Magnet, there is a 70% lift in mobile traffic percentage. It is critical for these campaigns that we design our conversion funnel with mobile users in mind if we want to turn Lead Magnet downloaders into purchasers.
So that’s how to dig into the cold hard facts about the audience, and we’ll use that information to shape the targeting of the campaign during our optimization phase.
But what about the messaging?
What about our follow up strategy?
To figure out our best tactics here, we’ll turn to psychographic data.
We’re going to create another custom report, this time looking at Affinity Categories and In-Market Segments. This will follow the exact same process as before except that our Dimension Drilldowns will be Affinity Category and In-Market Segment, respectively.
So we’ve got our report, now let’s look at the actual data.
When looking at these categories, a comparison to averages is what I like best because it highlights interests better than raw numbers.
I’ve pulled some of the favorite and least favorite likes of this audience, and the results are interesting.
The broad categories like “Movies” tend to ride towards the top, while categories that drill down tend to float towards the bottom. But, you can see that Entertainment & Celebrity News Junkies made it into the top 2 interest segments, which is huge.
An ad that asks “How does your business’ Social Media Score compare to Kim Kardashian’s?” would likely be a huge hit with this audience. On the other hand, sports-themed ads would most likely flop (spoiler alert – we ran some soccer themed ads and they did not perform well).
When you’re trying to think of messaging, focus on more niched-down categories than the broad high-level ones.
Sure, this audience likes movies, but they clearly HATE horror films. So, if you went in blind and tried a scary movie theme with your next ad campaign, there’s a good chance you’d have a flop on your hands.
While Affinity Categories look more at likes, In-Market Segments give us some indication as to what this audience is in the market to purchase (or may have just purchased).
I like to alphabetize In-Market Segments to look at interest clusters across category subsets. So here we can see that Education is interesting to the people who opted into our Lead Magnet, particularly childhood education and college and graduate studies.
Dating services is clearly what stands out, and is something that this target audience is highly interested in. So ads that used the language of education or dating are set up for success.
Lastly, we can use our sleuth powers to get a sense of whether or not this target market is interested in luxury goods, and therefore get a sense of whether or not they fall on the high end of the spectrum of income.
To do this, we’ll see if they have high concentrations of interest in categories that indicate wealth, namely luxury items.
Here are some general categories to use for this analysis:
Only about 3% of our audience for this campaign showed up with any of these segments, indicating that the people opting in for this lead magnet are not notably wealthy.
So, following up with a campaign for a particularly pricey item will likely be less successful than one with more moderate, tripwire-style offers.
Let’s combine all the different data points we got on our audience and figure out what the profile is of our ideal candidate for the 10-Minute Social Media Audit:
Age: Late 20s, early 30s
Preferred Device: Smartphone
Estimated Average Income: Under $100k Annually
Role at Work: Non-Supervisor
Likes & Interests: Movies, Celebrity & Entertainment News, Pop Music, Education, Career Consulting, Dating Services, Home Décor & Gardening Services
Dislikes: Sports, Horror Films, Boardgames, Videogames, Automobile Accessories
We’ve now constructed an ideal candidate profile for Facebook Ads we run to our 10-Minute Social Media Audit, based on what has worked so far. With this information we can attack the campaign from a new angle, and speak more directly to our target audience.
Imagine what your paid traffic team could do with this level of detail about whom they are targeting.
Don’t forget that this can be applied not just to people who opt-in but to purchasers or, with careful targeting, membership site users. In addition, you can use this strategy for any platform that drives sufficient traffic volume.
Have a question?
Ask the DM team and 9,227 other members in the DM Engage Facebook Group!
Not a DM Lab Member? Learn more here.
John Grimshaw is Digital Marketers Marketing Coordinator. He is the resident number-cruncher, analyzing data from emails, websites, and customer service. Armed with this information, he suggests how to improve both our promotions and the overall customer experience. Connect with John on Twitter.View all Posts by John Grimshaw