Tracking campaigns and website traffic
Curious about how to track your campaigns in Google Analytics the right way? I hear you.
Well I hope you will agree that tracking your campaigns is absolutely critical not only to your marketing department but to the overall performance of your business.
Tracking your campaigns helps you to understand what is working and what marketing tactics are a big fat flop. But it’s also important to use the data from your analytics to backup your budget and resources for all your super smashing marketing wizardry.
What if I was to tell you that you may be doing your tracking all wrong?
Would you feel nervous? A pang of panic?
Have you made the right decisions when sacking that agency for not delivering, or put money on a channel that was actually, well wrong……?
As a marketer (or anyone who’s job is to get people to your website) you have to really understand how Google Analytics works.How they define your website traffic, how they may be bucketing your wonderful work, which could be, for all your best intentions and tagging efforts being thrown in the wrong bucket, or sent to (none) / Direct which is what I like to call, “the-man-draw-of-data-hell”.
But fear not my friend, The Colouring In Department are here to save you!
In this post I am going to walk you through:
- How UTM tracking codes works
- How Google Analytics actually defines your marketing channels
- Common mistakes with UTM tracking and your sources and mediums
- What you can do to fix your tracking problems (hurrah!)
- How to tinker with your Default Channel Grouping settings (spoiler, you need to do this, especially if you are doing paid social)
How UTM tracking codes works.
You use UTM codes to track your marketing campaigns, as we will dig into the nitty gritty in this post you will see that GA needs some help in making sure your marketing gets put in the correct pot. A side note here, UTM codes are for your EXTERNAL marketing campaigns, you DO NOT, and I really mean DO NOT use UTM codes in your own website, you will royally fuck up your data. If you did want to know who went from the Blog to your Product page you can find those answers in the Behavior reports and by building Segments. Right, now that is out of the way……
Let’s break down a UTM code.
You need to have a URL that you want the traffic to go too, this sounds uber obvious, but you want to check that you have the correct URL, check it works (no one wants a 404 page when they click on that link), check its current eg your HTTPS version not the HHTP that will redirect.
This is the big broad buckets for your marketing channels, Organic is a medium, Email is a medium, Social is a medium.
This tells us where does the link live, so if Organic is the medium the source would be Bing or Google, if Email is the medium it will be the name of the data set eg newsletter database, or customer database, and if it was Social the source would be Facebook.com or Twitter.com
You can choose to give the link a campaign name, so if you were doing a marketing campaign for a promotion that included a bit of PPC, Social Media and Email you can name all your traffic involved in said promotion under one campaign name which you can dig into in your Analytics , which is a massive timesaver to see how the collective channels are working for a campaign.
Love using this 💕
It gives you the option to further slice your data. If you had an email campaign, you would set the Medium = email , Source = newsletter-database, Campaign = BlackFridaySale and if you had, say a CTA that links to the same URL as say an images, you could create 2 UTM codes, the only difference is to vary the Content tag, one would say Content = CTA the other Content = Image. This would give you insights into your email creative and how that is working at driving traffic and conversions.
Below is a good example of a correctly tagged email campaign. Medium is correct set as ’email’ , and the source is lovely, where does that link live…..it is from the ‘newsletter-database’, so all the traffic from this person’s email marketing is going into the right bucket and being attributed correctly. High Fives!
How Google Analytics actually defines your marketing channels.
So far, you may think, this all sounds simple doesn’t it, what could go wrong?
Well a lot.
You need to help Google Analytics understand where you’re inbound links are coming from to help attributes correctly. It is after all, a computer programme!
So you really need to know how GA processes that information and I know you guys do not lay awake at night thinking ‘I wonder how Google Analytics core reporting API defines the Default Channel Attribution model for my Acquisition reports”
Lucky for you, I have (true story) so here we go!
The Default Channel Grouping (which you can find in your Admin> View settings) is what powers your gorgeous Acquisition > Channels report. You know, the one you use to see how you marketing is working for you.
Think of the Default Channel Grouping as a bouncer, they are at the door of your website, asking people where they have come from and checking through their little list to see if your traffic matches the rules that they have down.
This is where the problems can start.
Let me backup and explain how Google defines this list and show you some common errors (that you may be doing and will need to fix).
Below is the actual list of rules for the Default Channel Groupings, this is the little list the bouncer is checking your traffic against, and it all HAS To match exactly what is here, case sensitive, no room for error.
1- Referral. This is a click through from any website that Google does not recognise as a search engine so this can come from other websites that maybe you have a Blog link on, social media is also included in this bucket.
2- Organic. These are Google recognised search engines so Google, Bing, Yahoo (side note, I find it rather amusing that DuckDuckGo which is a search engine is categorised as a referral medium by Google)
3- (None) This is a for your inbound links that are pushed directly into the browser and with no referral data or parameters to see where it came from, it gets recorded as source = (none) and therefore medium = (direct)
You will notice that there are a few more channels listed when you select Acquisition>Channels but again you have to understand that this channels report is powered by Google’s Default Channel Grouping, this is Google system defined channel definitions, and it is locked down.
This is how google buckets your hard work, notice that there are no channels for paid social, or retargeting! We will get to how you fix that in a second.
Google Analytics will go through its lists of SYSTEM DEFINED rules and based on that ,will assign the channel they will appear in which is what you see in Acquisition > Channels.
If you get it wrong, well your in for some bad luck.
Common mistakes with UTM tracking and your sources and mediums
Let’s start nice and easy with just using UTM codes and NOT matching them to the system defined. Email is the big common issue I see here.
If you look at the system defined rule the Medium = email , as in lower case email, not Email , not Email Marketing, not Newsletter. If you put something in a UTM that the bouncer does not understand it will be sent to (other) which is kind of where data goes to die and collect dust.
Then we have the problem when people do not even tag email marketing (tut tut to you!)
Let’s say you send out an email and 1000 people click through to your website. 50% of the emails were click on from people using Outlook on their desktop, and the other 50% through an online email client such as Gmail.
Without any campaign tags on those inbound links, Google Analytics is going to count the outlook link traffic as Direct because it won’t find any parameters so It will be treated as if it was Direct traffic, and if the link was opened in an email in Gmail browser, Google Analytics thinks that it is a website referring traffic to you so the medium assigned will be Referral and the Source = gmail.com or something to that tune.
Both are wrong, and Email gets no love (poor email marketing channel).
Here is another example of a bad Now if you look at again at our Default Channel Grouping rule list, there is NO medium for social-post so that is thrown in to (other) along with another set of email traffic as someone tagged the medium as browser, which also does not exist in the Default Channel Grouping.
And it came again for a client that didnt tag their email correctly.
Personally, I have done a few hundred analytics audits, and over the years have taught thousands of people about the joys of Google Analytics, and one of the biggest issues I see boils down to people not correctly tagging.
But do not fret dear reader, you can tinker with your Default Channel Grouping settings (spoiler, you need to do this, especially if you are doing paid social)
What you can do to fix your tracking problems (hurrah!)
Get a sense check to any issues you think you may have, and talk to your team about data discrepancies they think they have eg “We are doing loads of email campaigns but the data never shows in GA.”
This is what happened with this example below, they had about 3 members of staff full time on email, sending out around 5-25k emails a month, yet is looked like they were the worst department in the world.
They NEVER tagged the emails so it got thrown into Referral and Direct (sad times). Check out the image below. This was from their account…… when we dug into the data, the Referral traffic was full of source = gmail, yahoo, outlook, all that email traffic was being opened in an email browser and therefore not attributed correctly.
Another common mistake I see is with Paid Social Campaigns. Did you spot that the Default Channel Grouping has NO way to group or define paid social.
The problem with tagging paid social campaigns for say Facebook, if you have tagged the medium as CPC, the Default Channel Grouping has a System Defined channel called Paid Search and if the medium matches CPC it gets put in that bucked, AKA Adwords!
Now it’s not technically wrong, Facebook is on a cost per click basis, but boosting your social posts is a very different strategy and tactics than those in your search marketing and your AdWords campaigns.
I’ve seen people login to Analytics, go to the Acquisition>Channels, and they take this report by its face value and sack agencies managing paid social as the numbers do not show in Analytics, or pull budget lines for other channels, like Email, as it doesn’t pull in the numbers.
As a general rule then, you have to do EXACTLY what is on the tin, you can not create new mediums, decide you would rather put a capital on the name when its lower case sensitive.
For me this shows just how critical tagging is.
Everybody within the team and external (agencies) basically anybody that’s creating inbound and trackable UTM links for any of your marketing campaigns needs to understand how important this is.
Because you are making decisions based on how well the channels in the Default Channel Grouping are doing.
How to tinker with your Default Channel Grouping settings (spoiler, you need to do this, especially if you are doing paid social)
Step 1: Work out your medium lists.
Think of mediums as big broad buckets for you to put your marketing activity into.
In addition to the System Defined Channels, sit down with your teams/agencies and work through all the mediums you need or have up and running that may not be showing up in your analytics correctly.
Remember the examples from the start of the post? Paid Facebook campaigns tagged as CPC were put in the Adwords bucket (Paid Search), or Email campaigns that may come up as Direct or Referral.
Below are some examples listed in the tracking sheet that can help get the brainstorm juice flowing.
Do you have PDF documents that you send out that have links in them? Are you doing Paid Social? Retargeting on Paid Social? Do you have a list of partnered website that you work with or do co-authored content?When it comes to your naming conventions, there is no right way, but do remember that these medium and source names show up in the tagged url for your customers and prospects to see.
You want to make them as descriptive as possible as you need to understand them quickly in your analytics, if you were to move jobs or handover to someone, it should be clear what your mediums and sources are.
Step 2- Create your User Defined channels
When you have worked out what mediums you need, you’re going to have to tinker with your Admin settings in order to tell Google how you want your channels to show up in your Acquisition Reports.
Here are some examples of Default Channel Groupings that have some User Defined channels for the marketing teams to identify how well their marketing campaigns are doing and how their website traffic actually got there.
There are a few ways you can tackle this, you could create a Custom Channel Grouping (beta) but that means you have to build up the Channel Groupings, including Google’s system defined channels, from scratch which I personally find a big fat ball ache, and prone to error.
Instead, I want you to head over to your Admin> View> Channel Settings >Channel Grouping and select Default Channel Grouping.
VERY IMPORTANT : What you are going to do here is changing the way that Google crunches your data and it does this from the day that you create them. If you make a mistake you can flat line your data. So it is very VERY important that you do this in your Test View first. When you are happy with the data coming in, create the same changes to your Reporting View.
Just in case some of you are going “what you talking about Jill”, let me back up quickly and explain.
Google recommends that you have 3 views in your Google Analytics account:
- One Raw Data View, think of this as your backup file
- A Master View or Reporting View ,you can call it anything you want, this is the view you put filters on, set up goals etc
- A Test View is so you can check things work and don’t bugger up your data.
If you want to know more about what your account structure should look like, check out our explainer for getting your Admin Account Properties and Views in order!.
Identify how you will define your new channels in your tagging conventions. When you create your User Defined Channels you need to use and keep using (consistency is key here) how you are going to tag your mediums in your UTM links. Below are some examples of NEW channels you may want to create and how you may want to name them.
Let’s say I wanted to tag traffic from SMS messages. I can create a rule where the Medium = sms and every time I track a link from an SMS it will no longer go to Direct, but will be moved to my nice new channel and show up in the Acquisition Reports.
How do you do this? It takes a few jiffy seconds! Click on the ‘define a new channel’, and provide Google with the details on how they are going to process your super UTM and tracked traffic to your website.
Then scroll down to the bottom and hit save! Always scroll down and hit save. You are working in a browser and if you don’t do this, all your work will be gone.
Rinse and repeat for all your new channels.
You need to be aware that these channels run in order so you want to put your more specific channels at the top and you’re more generic channels at the bottom.
I have given some guidance here on how you may want to tag your mediums, I have used ‘medium matches regex’ so it is not case sensitive. Tip Brian Clifton has a super post on using Regular Expression.
Again, you are going to follow this step in a test view and then when you are happy, roll it over to the reporting views you are using.
Step 3- Agree on Source names
The source tag is easy to work out and define, as this is where the link lives.
But again, you need to have consistency across the board, you don’t want fragmented reporting due to inconsistent tagging, see the example below.
In this case, the marketing team were using ‘facebook’ and ‘facebook.com’ for the Paid Social campaigns.
Another question I get asked is what the source should be for Email. I personally like to put the source in relation to the data bucket.
For example I am sending out my newsletter, I would put this down as Medium = email and Source = Newsletter+Database
Or I am sending a triggered email campaign to people who downloaded content, I would put Medium = email and Source as content+trigger and the campaign name, something like: Q12017analytics+white+paper
Tip: adding a + in the middle as turns it into a space in your reports, much easier on the eye!
Campaign Tagging Structure
So you’ve learnt how to track your campaigns in Google Analytics, now what?
As mentioned earlier in the post, you will need to maintain a log and this your the bible!! This is the list of all your links which everyone in the team (internal and external) will use so that your website traffic goes into the right bucket. We have such a document if you would like it, head over to our Resources.
But in general this is what you are going to agree on, no more getting your sources and mediums mixed up no fragmenting data with people having 40 million different ways to say the link was on facebook. Nice and tidy structure to help that computer programme understand where the link came from.
|MEDIUM||utm_medium=||big buckets for marketing channels, defined by system and user defined|
|TRAFFIC SOURCE||utm_source=||where the link lives|
|AD CONTENT||utm_content=||more info to help slice and dice data eg banner , mpu, text links etc|
|KEYWORD||utm_term=||used for Adwords|
Step 4- How to Understand Attribution and Assisted Conversions for your User Defined and System Defined Channels
What is attribution?
Short answer, it is a fancy way of saying ‘who gets the credit’.
If you haven’t used attribution models in Google Analytics before it can be little confusing, daunting or a mixture of both. But what exactly is attribution modelling and why is it so important?
According to Google, attribution modeling is:
An attribution model is the rule, or set of rules, that determines how credit for sales and conversions is assigned to touchpoints in conversion paths. For example, the Last Interaction model in Analytics assigns 100% credit to the final touchpoints (i.e., clicks) that immediately precede sales or conversions. In contrast, the First Interaction model assigns 100% credit to touchpoints that initiate conversion paths.
In layman’s terms, attribution modeling allows teams to determine the value of their marketing channels that led to a conversion.
Everything north of the Multi-Channel Funnel Reports is pulling in data from the core reporting API, which works on the last click win rule.
When we marketing folk like to see what is assisting in conversions, all that lovely work we have just done works in our Acquisition>Channels Report, last click win. BUT the reports in the Multi-Channel Reports run on the MCF Channel Grouping.
So how do you see if your Paid Social for example, is assisting in conversion?
There is a very quick way to do this. In your Conversions> Multi-Channel Funnels> Assisted Conversion Report click on the drop down and select copy MCF channel grouping template.
Now rinse and repeat the same process that you have just done for your Default Channel Grouping User Defined channels.
Unlike the Default Channel Grouping, this MCF Grouping is part of your personal tools and assets, and is only changing the way your current, and future data is being displayed. That means if you create this, you need to share this asset with people in your team.
- It is vital for anyone responsible for driving traffic to your website to understand how Google Analytics processes your traffic.
- It is our responsibility to tell Google what we want to happen to our inbound links, do not expect them to put them in the right bucket.
- Use the option to add a Secondary Dimensions in your reports, in this case, adding Source/Medium to identify if you have any flags that need looking at urgently.
- Work out what channels you have, and need, that are not part of the Default Channel Grouping.
- Work together with your wider teams, agencies, anyone who is responsible for your marketing to get them to understand how you are going to Tag and Log your inbound links.
- Agree on a naming convention, consistency here is the key
- Keep a log of all the links you create: use our channel planning template as a starting point.
- When you are ready to tinker with your settings, do this in your Test View first then when you are happy, recreate in your Reporting View
- Don’t forget about Assisted Conversions, copy your MCF Grouping and repeat the steps in your Default View so you can see how your channels, including any new ones, are working at assisting conversions for you.
Hopefully by reading this far into the post you have seen the importance of getting your channels and tracking in order, but how do you do this?
This is your to-do list, you need to own this, consistency is key, so you want to bring in your team members, and brief your agencies/consultants so you are all on the same page.
And for the love of marketing, keep a log! I have seen people report fragmented sources or teams working in silo and tagging more than one different campaign the same name! (Rolls eyes).
To help you, we have created a UTM explainer that you are more than welcome to use.
Found this post interesting?
We have a whole module on Tracking in our Online Google Analytics Course.
And, as no one likes a blank sheet of paper, we’ll walk through a process to work out your channel planning process so you marketing goes into the right place
You know you want to have a look 👀