CXL Institute Digital Analytics Minidegree Review

You may not be there when all of this started, but then you got to know from where to pick and start. Noel Gomes, on Google Analytics Audit for Marketers.

As a marketer, you probably may come to the party late. It is quite likely that Google Analytics is already set, things are pretty much moving around, but somehow the one who hosts the party isn’t happy with the music, and isn’t happy the way things are going about. That’s where you come in.

Like most cops and detectives, as soon as you arrive at the scene what do you do? Analyze and audit. That’s what we are going to learn from this blog. How do you get started with an account that has been doing business for quite some time, and you are recently been handed the reins.

In this blog, we plan to cover the following points under Google Analytics Audit.

• Account and Property Overview
• GA Property and View
• Sending Page View Correctly
• The Hostname Filter
• IP Filters
• Default Channel Group

But before that a short prelude that introduces us to the tool we will require to conduct the Google Analytics Audit.

Things we are going to learn in this part of the blog

• How do you start to use Chrome Developer tools?
• Why and how to preserve navigation?
• Learn about Chrome extension and other tools for Google Analytics Audit

Why do you need to know how to use Chrome Developer tools?

With the Chrome Developer Tools, you can develop, test, and debug your websites right within your browser.

How do you start to use Chrome Developer tools?

There are two ways to do this.

• Windows — Press F12 Mac — fn +F12
• Windows — Ctrl+Shift+I Mac — Command(?)+Option+I

Chrome extension and other tools for Google Analytics Audit

The tools we will be using for Google Analytics Audit are:

• Google Tag Assistant
• Adswerve — dataLayer Inspector+
• Google Analytics Debugger
• GTM/GA Debug

Out of the above tools, the first two tools would be our most used ones. I have hyperlinked the names of the tools so that you read what these tools do before you install them.

Here’s a website, I recently visited to buy some merchandise. Let’s use the Adswerve — dataLayer Inspector+ to look at my journey through the website. To do that we will need to call the Chrome Developer tool. Command(?)+Option+I. Place the DevTool screen to the bottom half.

Screenshot #1: Chrome Developer Tools → Console>Preserve log upon navigation

Make sure the checkbox —’ Preserve log upon navigation’ is ticked. Once this is ticked every time you move within a website it will track the navigation using Adswerve — dataLayer Inspector+.

Screenshot #2: Inspector tracking navigation

The GTA/GA Debug and dataLayer Inspector+ pull out similar information. Just that the GTA/GA is visually better to look at.

Google Tag Assistant.

You need to activate the GTA to analyze all pages on the current tab.

Screenshot #3: Enable Google Tag Assistant

Enable the Google Tag Assistant and it will list out all the Google Tags on the current page in the Google Tag Assistant. Another important feature of GTA is the Record button, hit it and it will record an entire session and send you the details which you later can look back to see the kind of the information you received.

The First Part of the Blog:

Google Account and Property

Topics we would be covering in this part of the blog

• Document basic info about Account and Property
• What is sending info to Google Analytics?

a) Classic GA

b) Universal GA

c) Global

d) GTM

So let’s kickstart with Document basic info about Account and Property.

The things you will require to document basic information:

1. Google Analytics Audit Template
2. Google Analytics Demo Account
3. Tools: Google Tag Assistant, and Adswerve — dataLayer Inspector+

Screenshot #4: Google Analytics Audit Template

Pull out the Google Analytics Audit Template and get going with the template.

Screenshot #5: Account Name

Screenshot #6: Property Settings

Screenshot #7: View Name | Number of Views

Screenshot #8: View Name | Number of Views

For GA — which version (classic, universal, global — ga.js, analytics.js, gtag.js), we will come back to this later.

In the sheet, scroll down to the Account Level.

Screenshot #9: Filter Details | User Management

Screenshot #10: Filter Details | User Management

Inside Google Analytics Demo Account, you have by default, two filters — Rename Adwords Campaigns and Include Hostname. Also in User Management, it has just my name.

Screenshot #11: Update Google Analytics Audit Template

Three Examples of Sending Data to GA

1. Analytics.js code snippet (Universal)
2. Gtag.js code snippet(Global Tag)
3. Google Tag Manager

Screenshot #12 | Google Tag Assistant

In the above screenshot, you see there is a list of Tags that are tracking data on this site. Click on each of these tags to learn more about the details.

The GTA shows that the site is using Google Analytics. It also has the other GA tags. On clicking these you will get to know details like Account ID, Conversion ID, URLs, etc

Screenshot #13 | Google Analytics → Universal

Screenshot #14 | Global site tag(gtag.js) → Global Tag

In the list you see, the website is also using Google Analytics. If you go behind the page, into the page source you will find the Analytics.js snippet code(screenshot below).

Screenshot #15 | Analytics.js code snippet → Universal

Analytics.js is also known as “the Google Analytics tag.” It is a JavaScript library for measuring how users interact with your website.

To the end of the list, is the Google Tag Manager. It is a free tool that allows you to manage and deploy marketing tags that is snippets of code or tracking pixels, on your website/mobile app, without having to modify the code.

Screenshot #16 | Google Merchant Store → Google Analytics Tags

Let’s get back to our audit sheet. To fill in the remaining details run the Google Tag Assistant. Go to “GA — which version (classic, universal, global — ga.js, analytics.js, gtag.js).”

Based on the screenshot above, we see it has a Global Tag and no GTM tag.

While referring to the example of a furniture based website(screenshots #12,#13,#14) you see that all the information is going through GTM, using Universal GA.

Screenshot #17 | Filled the Overview Part of the Sheet

Second Part of the Blog:

Google Analytics — Property and Views

1. Introduction to our problem-ranking scale
2. Verify key portions of Property
3. Verify/setup appropriate views
4. Check key parameters of main view and raw view
5. Continue filling the spreadsheet

Problem-ranking Scale

The Problem-ranking Scale is based on the Eisenhower Matrix. It is also referred to as Important-Urgent Matrix. It helps you decide on and prioritize your task based on what is important and what is urgent. Once you list and sort tasks based on the least urgent and important tasks, you go up the list to see what gets listed in the most important and urgent tasks based on the degree of importance and urgency you decide which to delegate and which to skip.

The four parameters of the ranking scale are:

1. Important and Urgent
2. Important but Not Urgent
3. Not Important but Urgent
4. Not Important but Not Urgent

In our audit spreadsheet, we will mark the items based on points 1 and 2. Important and Urgent. Important but Not Urgent. We cannot rule out the use of the other two parameters.

Say, Not Important but Urgent. The client may like us to see into the Bounce Rate while we do the audit.

Screenshot #18 | Property and View Audit → Important | Urgent

We start filling in the details for Property and Views. See you inside Google Analytics Demo Account>Property.

Screenshot #19 | Default URL

The default URL is not always set up properly. One nice way to check that is using a chrome extension → Link Redirect Trace.

Screenshot #20 | Link Redirect Trace

SpaceX is one of my favorites websites, I come here regularly to shop for accessories and learn about updates. I chose this website as an example to check for the default URL. I typed spacex.com into the address bar, I then clicked on the Link Redirect Trace. It showed me that I was redirected to https://spacex.com/. which got me redirected to the https version of their URL https://spacex.com/

We are not auditing SpaceX here. If we were, we would suggest clipping the redirects to http: and redirect the users directly to the default URL https://spacex.com/. We would have set up the https:// as the default URL in the Property Settings. But for now will mark it as Important, but not Urgent.

Next on the sheet: Default View. Refer to screenshot #19, it shows 1. Master View as the default View. In the Property Setting scroll down to the part Search Console, try clicking it, what does it show?

Oops! there is no Search Console set up. Let’s look at a GA account that has the Google Search Console setup. How does it look like?

Screenshot #21 | Google Search Console

You will see the GSC uses a https://www as the default URL.

We now scroll down in the Property>Tracking Info>Referral Exclusion List. By default, you will have googlemerchandisestore.com. Ideally, in an active Google Analytics account, you will see a couple of domains listed here, first being its own domain name, the others would include the domain name of the payment gateway you installed on your website, also the domain name of your CRM, emailing software or webinar domain name.

The next important thing we need to look at during the audit is the Custom Dimensions. Google Analytics Demo Account>Property>Custom Definitions>Custom Dimensions.

Screenshot #22 | Custom Dimensions

That’s the default Custom Dimensions in a Demo Account. In Simo Ahava’s Blog, there are some good suggestions on setting up 13 Useful Custom Dimensions for Google Analytics. I am already setting up most of these for my clients.

Screenshot #23 | Top 6 Custom Dimensions

In the screenshot above, a look at the Custom Dimensions of a client’s GA account. Most experts recommend having these six Custom Dimensions. The most important ones are in the red box. Client ID and Hit Timestamp. A great help in debugging issues with GA.

Going back to the Audit Sheet>Views. While learning to setup Account, Property and Views in the first blog, we had already created 3 views → 1. Master View 2. Test View 3. Raw Data View.

Screenshot #24 | Audit Sheet>Views

Next on the sheet is the Main View. Go to your Google Analytics Demo Account. First, check if the website URL in the Main is the same as that in the Property Setting. Second, make sure the Main View>Bot Filter is checked. Little further, you will see if the Linked Google Ads Account is linked. Also, check if the Site Search Tracking is on and they are working properly.

Going by the audit sheet, next we have is the Annotations.

Screenshot #25 | Annotations

These are the annotations available by default. An account using annotation is a great sign, it means someone is paying attention to the Property.

Check if there are any Custom Alerts. The demo account doesn’t have one.

The one I would recommend is creating Custom Alert → No Traffic.

Screenshot #26 | Custom Alert

Here’s a custom alert that tells me when the session is less than a value, in this example say, less than two. Let’s go back to the audit sheet and fill in the details. The demo account by default doesn’t have any custom alert set up. Setting up an alert around low traffic can be very important to understand if there is any slump in the traffic. And so, it is Important but Not Urgent.

Screenshot #27 | Audit Sheet → Main View

The next part of the audit sheet → Raw View. In the Demo Account, 3. Raw Data View. By default, you have 2 filters. However, considering the importance of Raw Data we suggest not to use filters in the Raw Data View.

Screenshot 28 | Raw Data View>Filter

Screenshot #29 | Raw Data View

Some of the BIG NOs and a BIG YES for Raw Data View Settings. You don’t want the data in the Raw View to be filtered.

We now move to 3.Raw Data View>Ecommerce Settings. It is rather useful to have the Enhance Ecommerce Setting in the Raw Data View as same as the Master View.

Ideally, the good practice for setting up views is to have 3 views → The Master View, The Test View, and The Raw Data View. All of this varies from business to business, what if your site gets equally good traffic for organic and paid. In such cases, having a view for organic traffic and for paid traffic would make a great idea. In the Organic View, we would set up a filter with the filter pattern as Organic. For Paid View, we would set up a regex something like → ^(cpc|ppc|paidsearch|display|spm|banner)$

Let’s fill up the audit sheet for Raw View. As the Raw Data View has filters we would suggest to lose them, we would like the data to come in as it is, without any filter. The use of filters may lead to loosing on to something we wouldn’t want to take our chances with. Hence, I have marked it Important and Urgent.

Screenshot #30 | Raw Data View

The Enhanced Ecommerce View is all funky. We will need to work on it to get useful e-commerce data. For now, I have marked it as Important, but Not Urgent.

Third Part of the blog:

Sending Page View Correctly

Google Analytics Audit Sheet> Accuracy → Page View.

Under this, we will be verifying Page Views that are being tracked. Are there instances of undercounting or overcounting? However, we will not talk about GDPA/CCPA, these can legitimately block page views. We want to make sure that the Page Views are sent to the GA properly.

Let’s take an example to verify that the Page View is sending data to GA. Let’s get on monstertore.com. They are into making cables for guitars and PAs. Check what does the Google Tag Manager pulls out.

Screenshot 31 | Google Tag Manager

We see two GTMs deployed, a Global Tag, and Google Analytics. The site is using GTM means the Page Views are being tracked. Let’s pull out the console to see the tracking.

Screenshot #32 | GTA Page View

The console has a different story to tell, of the two GTM containers, used, we see only one working, on zooming in we see that this page view has not been sent by GTM. We can do one thing to see if the dataLayer is tracking any data. We go to the menu of the website and go to CABLES.

Screenshot #33 | Page View Tracking

No page view came up. Let’s try to add-to-cart.

Screenshot #34 | EC, EA, EL

Strange. The Event Category shows Cables, Event Action shows Products, but the Event Label says, unidentified. Ideally, you include product as the label, not an event action. This happened not once, twice or thrice, but four times and on clicking them, it says, not sent by GTM. There could be some duplication happening here.

If I were auditing this site, I would say undercounting, yes, and would certainly mark it as Important and Urgent.

Screenshot #35 | Google Tag Assistant

Screenshot #36 | GA Page View

Pull up the console and you will see it has GA Page Views. On clicking you see.

Screenshot #37 | GA Page View

The page view says it was not sent by Google Tag Manager. The one below says this is a duplicate page view. With such instances, our page views will certainly go wrong.

Screenshot #38 | GA Page View

If you are wondering why is this happening, the most common reason is that Google Analytics on the site was before and then GTM was added, without removing the analytics code.

Screenshot #39 | Google Analytics & Google Tag Manager → Source Code

To remove the duplicate page view on the site, we need to get rid of the Google Analytics code from the source code. If we were to audit this site, we would say, yes there is overcounting happening as the Page Views are not tracking correctly. There is duplication happening and there is a need to remove the analytics snippet.

Screenshot #40 | Google Analytics

The fourth part of the blog:

The Hostname Filter

Google Analytics Audit Sheet> Hostname Filter

We go back to patagonia.com, you know the drill → pull out the console. This time we use the Chrome tool → GTM/GA Debug. It has a neat way of presenting information.

Screenshot #41 | Console →GTM/GA Debug

Screenshot #42| Console → GA

In the GA information, we are looking for the hostname that is implicit of document location.

Screenshot #43 | GA>Document Location

To the right of the GA screen, you will find dl(document location), that’s the hostname. In case you are not sure what dl stands for, on the same screen, under Options, you see Parse Keys. This works as the dictionary or glossary for the terms you see in blue, listed under Object.

Screenshot #44 | Document Location>Parse Key

This is how the GA Screen looks like:

Screenshot #45 | GA → Page View & Events

Screenshot #46 GA → Events

The screenshots above are the left and right sides of the screen of the GTM/GA Debug tool. In the red box, you see the details of event action, event category, and event label. Again, in case you are not sure what the letters in blue stand for, hit the Parse Key to see what it is.

What makes Hostname special?

We see that valid GA hits include our hostname and if we were to create a filter that was specifically looking for hostname, we would be reporting on real traffic that comes to our website.

The tricky part is getting all hostname. It is simple if you have just the domainname.com and you don’t have the property ID, another sister website, or app defined inside your Property.

It gets tricky when

• you are doing cross-domain tracking
• you may have marketing automation pages
• you are having a landing page on a different domain

If you are having a landing page on a different domain, you surely would like to create a Property ID for it and track the data.

Remember: to verify the hostnames of the property that you are auditing, you need to look back for a long period of time, say a year.

See you inside the Raw Data View → Audience>Technology>Network.

Screenshot #47 | Hostname

Right from the beginning, the idea of having a filter in Raw Data is something, a BIG NO NO. Look at the screenshots below, they will support my belief and practice.

Screenshot #48 (left) Raw Data>Hostname>User | Screenshot #49 (right) Master View>Hostname>User

Screenshot #49 Master View>Hostname>User

Take a look at the number of users for hostnames in both views. It is 12,392. This is pretty unusual that both Raw Data and Master view have the exact same number of users. This goes to say that the Raw Data is not working properly.

That was something we saw in the demo account. Follow me, here’s the view to northwoodsoft’s → hostnames in a filtered and unfiltered view.

Screenshot #50 | Hostnames → Unfiltered and Filtered View

The filtered list is much smaller than the unfiltered. Hostnames → www.northwoodsoft.com, pages.services are my Market Automation tools. try.northwoodsoft.com is one of our landing pages and Googleweblight.com is used for slower mobile connections.

Tag along with me as we look inside the hostname filter.

Screenshot #51 | Filter View → Hostname

They have used a regular expression in the Filter Pattern

^(.*\.)?northwoodsofy\.com|pages\.services|cst3\.marketingautomat

The above regular expression looks out for 3 things northwoodsoft.com, page.services, and marketingautomation sites.

This filter ensures that only the selected domains are to be tracked and sent to the reports. So, let’s fill in the audit sheet.

Screenshot #52 | Audit Sheet → Hostname

Yes, hostname filters exist in the Demo Account. Apparently, it was also applied to the Raw View, we won’t say it is correct as we have already seen why in screenshot #48 and #49.

Let's check for Spam.

Remember: a valid hostname will block ghost referral spam, spam that is sent in through the Google measurement protocol.

How or when do you know your website is attracting spam?

Here’s a point in case. This client raised a flag when they saw this:

Screenshot #53 | Spam

“/legal.php → what is this?” The client said we don’t have any page with this title. And why is it getting so much traffic?

On looking at the numbers, you will see the number of new sessions is 100%, and the bounce rate is high.

Screenshot #54 | Spam Check

So we tried looking into the Service Provider and here is what we got. Now, this is unusual, “microsoft corporation.” There is something wrong here. Let’s check for other service providers, using Advance Filter.

Screenshot #55 | Spam Check Using Advance Search

In Advanced Search, we are looking for Service Providers whose Bounce Rate is Greater than 95 and Users Greater than 100.

The results were weird.

Screenshots #56 | Other Service Providers

You will see amazon technologies, microsoft corporation, hubspot, etc. which simply doesn't make sense. This surely must be coming in from some bot or is spam.

Please note: As of Feb 4, 2020, Google stopped reporting the Service Provider.

There is one resource I refer to

• Carlos Escalera Alonso’s blog on How to Efficiently Filter Spam, Bots, & Other Junk Traffic in Google Analytics.
• The Ultimate Guide to Stopping Junk Traffic in Google Analytics

Let’s fill in the audit sheet.

Screenshot #57 | Audit Sheet → Hostname Filter

The hostname filter as we saw, does exist and is also applied to Raw filter which we are not much of an evangelist. So have made a note that it needs to be removed. In this context as we are dealing with Raw Data View, I have marked it Important and Urgent to fix this.

Any indication of Spam? In reference to the client’s instance, yes we saw spam inflating numbers for the landing page. However, in the case of the Demo Account, our answer would be No.

The fifth part of the blog:

Internal Traffic Filter → IP Filter.

In this part of the blog, we will explore all the ways IP filters can be set up wrong. Although there are limitless ways it can be set up wrongly. But here, we will discuss a couple of them that we have come across.

Things we will do and use to understand Internal Traffic Filter.

• Use → Tag Assistant Recording
• Verify other exclusionary filters, to see if that's working
• Use the coffee website, the one we used in screenshot #34, #35, #36
• Use the coffee website views → Master, Testing, Raw Data
• IP address in the screenshot below
• Google Analytics Audit sheet> Other Filters

Screenshot #58 | (left) Coffee Views | Screenshot #59 (right) My IP

Screenshot #58 | (left) Coffee Views | Screenshot #59 (right) My IP

Screenshot #60 | Coffee Main View> Filter>Exclude Traffic

Let’s dive in. Inside filters you see exclude traffic filter. This filter is to exclude traffic coming from the IP → Screenshot #57. That’s the northwoodsoft.com website, we are referring to verify IP filter.

Screenshot #61 | IP address

Switch tab. Pull out the Audit sheet, under Other Filters>IP Filters Exists → we say YES. In the screenshot #60, we see this filter for IP address, now the question is does it work? Let’s check that.

Screenshot #62 | Google Tag Assistant

Pull out the Google Tag Assistant. You don’t have to enable this every time. Reach out to the three dots. Click Auto Validation to turn it ON, hit record, and go visiting a couple of pages. Now stop recording. Hit Google Analytics on the list of tags in the Google Assistant Tag and you will reach this screen.

Screenshot #63 | Google Tag Assistant>Google Analytics Report

In this report, the part we are interested in is the FLOW.

Screenshot #64 | Google Tag Assistant Tag>Google Analytics>Flow

Flow shows us the entire flow of my sessions from the start through the pages I visited up to the end of the session. At the top left, you see ‘change of location’? Click that, this helps you to check whether the IP address filter works. We will use a specific IP address, the IP address for Northwoods, refer to screenshot #58, and hit update.

Screenshot #65 | Google Tag Assistant> Change Location>Specific IP address

The Tag Assistant does something amazing. Let’s go to the FLOW.

Screenshot #66 | Google Tag Assitant>Google Analytics>Flow>Page View

After updating the ‘change location,’ we go to FLOW to see if the filter worked. Scroll up to screenshot #61, you will see that we were on the home page and from there we went visiting a couple of pages.

Cut to, screenshot #66. Under FLOW>Home>Page Views>View: Coffee Main → it says the hit was dropped by filter Exclude NSW traffic Shorewood IP. Refer to screenshot #59 & #60. Wow! this is great, Google Tag Assistant made it look so easy, and yes this means the filter is working. You can go inside the Audit excel sheet and answer the question → Are the IP filters working correctly?

Situation 1

Screenshot #67 | IP address

Check the IP address → 74.186.* do you think the filter will work? This filter is supposed to exclude traffic coming from the IP address mentioned in the IP address field.

Shall we verify? Let’s check. Go to the Coffee Website. Visit a couple of pages on the website. The Google Tag Manager will record these navigations.

Screenshot #68 | Google Tag Assistant>Google Analytics>Flow

Come inside →Google Tag Manager>Google Analytics>Flow. It says, ‘HIt captured without modifications.’ The filter is literally looking for “74.186.*, the real IP address of 74.186.777.333 does not match.

Remember: the IP address field is not a RegEx field. If your IP address is 74.87.71.66, do not try “74.87.*” It will not work.

Situation 2

Screenshot #69 | IP address

The IP address field reads as 74.187.71.0/22. Do you think a filter can detect this?

Well, there are websites that can decode it for us. This type of IP address is called a CIDR, i.e. Classless Inter-Domain Routing.

Screenshot #70 | Decoding the IP address

Copy-paste the IP address here, it will convert a CIDR into an IP address. Hit Calculate and you land on this screen.

Screenshot #71 | CIDR to IP Conversion

This conversion tells us that there is a whole lot of IP addresses blocked. Starting from IP:74.187.68.0 to the Last IP: 74.187.71.255. So there is a range of IP addresses that are blocked, then how do you check if the filter for these IPs is working? What type of filter do you put in GA?

There are two ways of doing this. One, create four filters. Two, create a filter using Regular Expression. The first way → create 4 filters, will cover four major blocks of IP addresses: 74.187.68, 69,70,71.

Screenshot #72 |Filter for a range of IP addresses

The filter is named, Exclude 74.187.68.x what it will do is look for IP addresses that start from this number and filter them from our view.

We’ve chosen Filter Tags>Predefined>Enable>Traffic from the IP addresses>that begins with and in the IP address field → ‘74.187.68.’ Notice, I have used a dot after 68. This field does not support regular expressions and so we haven’t used ‘x’ like the one in the Filter Name. This field takes the IP address mentioned here literally.

Create 4 filters like these the next for 69, 70, and 71. Hit Save, and you will be saved from unwanted tracking.

Another way is creating a filter using a Regular Expression.

Screenshot #73 | Creating Exclusion Filter Using Regular Expression

While in the previous exclusion filter, we chose to go with Filter Type>Predefined, however, to create a regular expression we will have to choose Custom.

In the previous exclusion filter, we had to enter an IP address, in the IP address field, this field does not support regular expressions. Here, in Filter Pattern, we can use regular expressions to filter out IP addresses we wouldn’t want to see in our data.

Filter Pattern: ^74\.187\.(68|69|70|71)\..*$ means to filter out IP addresses that start from 74.187.68–71. Now, let’s check if this works.

Screenshot #74 | Use a specific IP address

Pull out the Google Tag Assistant

1. Go to Google Analytics
2. Click Change Location
3. Use a specific IP address

In the IP address, give a random number in the fourth spot of the IP address, 74.187.68.143 (here, 143 is that random number). Hit update.

Screenshot #75 | Verifying the Exclusion Filter

Refresh. Let’s look inside the Flow. View — Coffee-Main: Hit dropped by filter Exclude 74.187.68.x

This filter works! Pause.

Go to the Audit sheet and answer the question: are the IP filters working correctly?

Anonymize IP

What happens when you anonymized your IP address?

Our IP address: 74.81.71.66, when we anonymize it the last octet, in this case, 66, will be dropped and usually becomes zero. i.e. 74.81.71.0.

We just learned in the above examples how Exclusion Filters work. Now if your IP address is 74.81.71.66 and you have anonymized it, so it now becomes 74.81.71.0, the exclusion filter will not work.

An anonymized IP can be set in G-Tag and Google Tag Manager. While running an audit you will have to check if there is an IP address that has been anonymized. Important. Important. Important, if there is an IP address anonymized your filter may break.

Situation 3

Screenshot #76 | Will this filter work?

Will this filter work? Filter Pattern → Madison Wisconsin. The Filter Name is Exclude Madison, we are excluding the city of Madison, Wisconsin.

This filter will not work, as the city is Madison and not ‘Madison, Wisconsin’. With this example you see, it is not IP filters that you can use. You can also use filters to include or exclude → for city, country, and region.

While you audit and you come across such filters, test it, go over to the excel sheet, and fill in the details. Like the one here below.

Screenshot #77 | Google Analytics Audit Sheet>Other Filters

The sixth part of the blog:

Default Channel Group

In this part of the blog, we are going to look at the next point in the Google Analytics Audit Sheet → Data Sources.

We need to make sure that the traffic from a different source and medium are bucketed properly. We will learn here, why do we need to do that, how to do that, and as we work through the Google Analytics audit of data source, you learn that this can be the most painful part of the audit.

Where in Google Analytics do you start the Data Source audit process?

Google Analytics Demo Account>Master View>Acquisition>All Traffic>Channels. Come here and you will learn → how people found their way to the website.

Screenshot #78 | Google Analytics Demo Account> Acquisition>All Traffic>Channels → Channel Grouping

This is the default Channel Grouping, it looks pretty neat. With almost 60% organic search and close to 20% from direct and a fair amount coming in from referral, things here look good seems good.

Let me take you to an active Google Analytics account.

Screenshot #79 | Examples of Channel Grouping

Screenshot #80 | Examples of Channel Grouping

Observe Other :) in screenshot #76, 75% of traffic comes from Other, in #77, almost 10% is from Other. In both cases, such instances showing up in the report are not a good indication.

Look at the numbers here, what do you think → what determines, which traffic shows up where?

You can reach out to the Help Center to learn how analytics classifies your traffic. You can check out the Default Channel Definition chart below:

Screenshot #81 | Channel Definition

A very important point, in the above table, scroll down to the last point. (unavailable) or (other), if sessions don’t match any channel description then it will fall under → Other.

Refer screenshot #76 & #77, Other → includes traffic coming from Medium or Referral that Google does not recognize.

Screenshot #82 When your definitions don’t match the default channel definition

Screenshot #83 | When your definitions don’t match the default channel definition

This is what happens when you come up with your own definition. THUMB RULE → FOLLOW THE DEFAULT CHANNEL DEFINITION. Going by the thumb rule, fb_hd, fb_carousel ideally belongs to the Social channel.

Follow the thumb rule or find such data tracked under OTHER. Remember, this is the traffic you are paying for, so why lose it to something undefined. Such data could fall under social or paid.

Rules or good traffic bucketization

• Read and know the Google Default Definition
• Adjust the tracking to the Google Default Definitions

Example: if you are doing cpc, look up the default channel definition, see what would this fall under? Paid Search? Then define it as Paid Search

• And/OR, adjust the default channel group to reflect the medium that you are using.

Remember: Every time you lose the opportunity to group your traffic under defined channels, it will show up as direct.

Going back to the example, refer screenshots #79 and #80. To fix this in 3 steps,

Step 1: Test View>Channel Settings>Channel Grouping

Screenshot #84 | Channel Grouping

Step 2: Click on, “Default Channel Grouping,” it will take you to the default setting.

Screenshot #85 | Channel Grouping Default Setting

Step 3: Inside the default setting, let’s click on Social.

Screenshot #86 | Add fb_carousel

Step 4: Add to medium>exactly matches> ‘fb_carousel’. Once you’ve made this change you will see this:

Screenshot #87 | System and User-defined

The status of channel>social changes to ‘System and User defined.’ We have made the adjustments in Test View, once these default channel settings kick in you will see it work while reporting inside test view. Apply to Master View when all clear.

Remember: always, always, while naming, renaming, or adding new names inside Google Analytics use lower case.

Email

Another area that can get messed up

Because by default most traffic from email are going to show as direct. How does this work? When you send emails to your audiences, and when they click on the link in the email, it will show up as direct.

Screenshot #88 | Campaign URL Builder

Inside the Campaign URL Builder setup the UTM parameters. Share the campaign URL (red box) in the email. This makes it easy to track the traffic coming in from email.

If you send a link in the email without setting up the UTM parameters all of that traffic will show up as (direct)/(none).

Remember: the medium and the source in UTM is always in lowercase as we have in the filter set the medium in lowercase.

ProTip: Have a medium filter set up in your property.

Let’s talk about how traffic can easily show up as Direct. What do you mean by direct? It is when someone typed in the URL directly or has bookmarked your website.

Top 4 ways when your traffic will show up in Direct.

• When you are on an HTTPS site and going to an HTTP site
• When there are too many redirects
• When you send an email with a link without UTM tag
• When you have links in PDFs and Word doc that are without UTM tags

Check out this blog that talks about the different ways traffic can show up in Direct.

It’s time to pull out our audit sheet.

Screenshot #89 | Audit Sheet> Data Sources

We need to set up an alarm the minute there is incorrect data source reporting. The minute the Default Channel Group — Other and Direct jump over the 20% mark, there goes the troubleshooting flag. We will need to pull out our lens to spot the error or problem.

For the Default Channel Group — Email, as long as you are using UTM tags, you need not worry. However, you can set a mark for a low percentage.

For the Default Channel Group — Other — Example Medium. Check if they are following the channel definition, refer to the example where we added fb_carousel as a medium under Social.

Check for the lowercase medium filter, if they don’t have, then set that up.

Well, that’s it for now, watch out for the continuation of the blog in Google Analytics Audit for Marketers — Part 2.

In this blog, we covered the following points under Google Analytics Audit.

• Account and Property Overview
• GA Property and View
• Sending Page View Correctly
• The Hostname Filter
• IP Filters
• Default Channel Group

Up to this point, we have understood what an audit can do to the process. Most marketers undertake Google Analytics Audit when they take over the system from its predecessor or from the client.

In part 2 of Google Analytics Audit for Marketers, I plan to cover the following topics:

• Site Crawl
• Content Grouping and Query Parameters
• Google Analytics — Events
• Google Analytics — Goals
• Personal Identifiable Information
• Enhanced Ecommerce

Get your Audit Sheets along see you in part two of the blog.