If you want to find out how to troubleshoot Google Analytics to improve your marketing efforts, this blog post is for you.
You’ll learn about the problems with GA that marketers frequently overlook, and how you can solve these problems.
Not only will I show you how to self-audit your Google Analytics, I’ll also give you a handy GA Self-Audit checklist at the end of this post. So you get a detailed blog post and a practical checklist for free. Because I’m nice like that. You’re probably one of the many internet marketers that use GA, but you don’t use it to its maximum potential.
Read on if you want to find out how to use Google Analytics more intelligently, by identifying the analytics issues that often lead marketers to make misinformed decisions.
A Strategy For Troubleshooting Your Google Analytics.
Generally, when preparing to audit Google Analytics to improve your marketing efforts, you’ll need to provide detailed answers to 4 key strategic questions.
Is your analytics tool properly configured?
Are you able to effectively track your users and their behavior?
Is your data accurate?
Can you track any additional items that you are presently not tracking?
Is your analytics tool properly configured?
Figuring out how to optimally configure the Google Analytics tool to give you the most useful measure of your website’s performance is an important first step.
Configuring the tool properly is all about getting the right data, and setting the right goals.
Rule of Thumb: Meaningful Data
A good rule of thumb here is that the tool should be set up to help you take actionable steps that improve your marketing efforts, based on the data that Google Analytics collects.
A great way to make your data more meaningful is to use filters.
Filters give you customized views of the data in your reports, instead of a broad overview that isn’t specific enough to base marketing decisions on.
Some examples of useful filters are:
- A spam filter
- An internal traffic filter
- Filter by subdomain
- Filter by web or application data
Let’s look at the obvious example of the spam filter. Just viewing your data as Google Analytics presents it (unfiltered) might lead you to make all sorts of incorrect conclusions about the volume of traffic you’re receiving.
But filtering out spammy visits could paint an entirely different picture of your site that will lead you to take steps to improve your SEO, rather than leave your site as it is.
Never delete or add to the original unfiltered view that Google Analytics generates for your site. If you want to use filters, always create a copy of your original view in order to preserve the original data. Otherwise, you’ll lose everything from when you first started tracking data with GA.
We all have our unique goals in life. Mine happens to be
world domination helping internet marketers like you kick-ass and earn money.
In a Google Analytics context, using goals provides a way to measure your website’s marketing performance.They represent the KPI’s that determine how successful your business is.
Firstly, make sure your goals are named intuitively, so you can easily identify what goal it is that you’re tracking. This seems obvious, but it’s easy to overlook. You’re likely to end up confused if you set arbitrary names such as Goal1, Goal2 etc.
A sensible goal name.
The most important thing about goals is to set intelligent goals. You can minimize superfluous information by sticking only to goals that measure how your business is performing. From a marketing perspective, this means focusing on conversions.
You should be tracking metrics such as:
- Number of conversions (newsletter signups, PDF downloads)
- Conversion rates for each marketing campaign
- Funnel drop-off points
The purpose of any goal is to track the actions users take to help you improve the bottom line of your business. If one of your planned or existing goals doesn’t meet this criterion, disregard it.
A visual representation of goals.
Out of the four goal types in Google Analytics, Event is the most useful in terms of tracking the number of user interactions with your site. But remember that if you want to track specific user interactions, you’ll need to modify your tracking code.
Assuming that you’ve installed the ga.js tracking code snippet on your pages, you can attach a call method to the particular element you want to track (such as a file download).
Event tracking setup.
A great way to get the most out of goals is to use a handy little setting called Goal Value. This assigns a monetary value to each conversion, which is brilliant for breaking down your revenue by marketing campaign or URL referral. This field is marked as optional in GA, which might’ve convinced you that it’s not that useful. That couldn’t be further from the truth, though. This is the sort of data you need.
If you are running an eCommerce site, you’re better off using eCommerce tracking in addition to goals, not just one or the other. Ecommerce tracking reports give you more insight into your revenue sources and your marketing campaigns, allowing you to see exactly which campaigns have lead to sales of particular products.
To enable eCommerce tracking
Go to property → view → select eCommerce settings → click the enable eCommerce toggle ON.
You’ll also need to make some modifications to the basic tracking code of your website’s pages in order to collect eCommerce data.
Ecommerce tracking code modifications.
Be careful when using third-party gateways such as PayPal on your eCommerce site. Google Analytics often gives credit to the payment gateway as opposed to the original traffic source when it analyses referrals, which can give you distorted data. This can be remedied by ensuring that the user returns to your website post payment and using the referral exclusion tool in the Google Analytics administrative interface.
Check below what else you can do
While it requires a more extensive code update, you can also get more bang out of your ecommerce tracking by using Google’s enhanced ecommerce tracking code. While the standard ecommerce tracking allows for data regarding transactions, product sales, revenue, tax, etc, the enhanced ecommerce tracking implementation provides significantly more in depth data on how users purchase and where they are falling out of the process. With enhanced ecommerce implemented, we can see user behavior from the site visit to the initial product impression, to when the product is added to cart, to a user viewing the checkout page, and all the way to the eventual purchase. We can also get more granular data at the product level, seeing how often a product is viewed and how that relates to “add to carts” and purchases.
Are you able to effectively track your users and their behavior?
Being able to track your users effectively is the big picture when it comes to Google Analytics.
In the context of GA, effective tracking:
- Enables you to identify which pages are hot with visitors to your site.
- View which pages visitors leave quicker than the time it takes Usain Bolt to sprint 100 meters.
- Get a sense of how your users behave on your site.
So how do you effectively track users and their behavior in Google Analytics?
– Check your tracking code.
The key point with tracking code is that it needs to be omnipresent on your site.
If you place tracking code snippets on some pages, but forget to do it on others, your analytics data will be incomplete.
Also, ensure that Google Analytics code is firing a pageview on every page, but only one pageview per page. Multiple instances of tracking code can skew pageview count (artificially high) and bounce rate (artificially low).
How To Check Your Tracking Code?
The first step is to ensure that data is being collected for analytics purposes. Thankfully, it’s pretty easy to rapidly check this. As soon as you add your tracking code, an automatic unfiltered reporting view is created.
Open up this view and go to Reporting->Real Time->Overview. If you see data here, your tracking code is functioning.
The second step is where things get tricky. You need to verify that the code is omnipresent on your site. This can be a pain in the ass if your website has hundreds or thousands of pages. Trawling through an entire site just to check if a snippet of code exists on each page is a painstaking process. A great website that speeds this up is http://www.gachecker.com/
This incredibly useful site analyses all of your web pages for that all important ga.js tracking code, saving you a lot of time and hassle.
You can also use Google’s nifty Tag Assistant Chrome browser add-on. If you have inserted your GA code using Google Tag Manager, this add-on allows you to check at the click of a button which pages have the Google Analytics tag. This can be extremely helpful for small to medium sized websites.
Google Tag Assistant Extension
If you want a more premium option, you can use the Screaming Frog SEO Spider Tool. The paid version of this software allows you to search for custom source code across your entire site, regardless of how many pages are on it.
So you can search for:
- Analytics.js for the presence of Universal Analytics code
- Ga.js for standard Google Analytics code
Note that in cases of websites using Google Tag Manager for analytics, there is no clearly defined way to check for the presence of tracking code with this, or any other tool. The best you can do is check GTM is actually firing on your site with the aforementioned Tag Assistant Chrome browser add-on.
Another convenient tool you can use is Visual SEO Studio. The program allows you to perform an audit of your technical SEO at a glance, including checking your tracking code. There is a free version and a premium version available.
With Visual SEO Studio you can:
- Crawl your site’s pages for the presence of tracking code (500 pages max on the free version, 150,000 on the premium version)
- Check for pages with duplicate tracking code snippets
The premium version is worth it if your site is sufficiently large.
Is your data accurate?
This is clearly quite important, seeing as you’ll be basing your analysis and actionable steps from the data you receive. If this data is shitty, you’re more likely to make misinformed marketing decisions.
How to recognize bad data.
Recognizing the causes of bad data is something that comes with experience. Here are a few good tips.
- Check your referral links. Often, an inaccurate data point is caused by phony traffic. You might think nothing of the fact that your site is suddenly receiving loads of referrals from some random website that you’ve never heard of. This is not normal. Check it out – click the link. If it’s in a different language, or you get redirected to another site, chances are that any traffic arising from here is bad data.
**This is taken from here
- Look at traffic patterns. A sudden spike in referrals from a certain domain, followed by an immediate drop to zero traffic is a big warning sign that there’s something strange going on. Traffic tends to peter out gradually – it shouldn’t just drop to zero.
Strange traffic patterns like this signify bad data.
- Missing data. This relates to the omnipresence of your GA tracking code. If the code is absent on one or more pages of your site, the integrity of your data will be compromised. Use the aforementioned methods to double check for the presence of tracking code.
- Use proper tracking structure. Many marketers place the tracking code just before the </body> tag out of worry that it affects page loading speed. They want the visual elements to take priority. GA tracking code is asynchronous, though.
Be careful with sampling errors!! When you segment your audience in GA and create custom reports for your marketing campaigns, the tool will create these reports based on samples. This a problem for websites with a large number of sessions. As your site is scaled upwards, you’ll encounter this issue more frequently.
Let’s say you want to develop reports to find out the conversion rate for marketing campaign Y, which has 2,000 sessions. When your total website sessions are > 500,000, GA will automatically apply sampling to the data. Let’s say you’ve got 20 million sessions, and GA uses 500,000 random sessions to create your report for marketing campaign Y.
The highlighted notification tells you when sampling has taken effect.
This report might estimate your conversion rate for campaign Y at 3%. But estimates based on a sample will likely differ from the actual value. This is known in statistics as the sampling error. The smaller the segment is, the larger the sampling error.
How can you overcome sampling issues?
- Increase the sample size.Google Analytics uses sampling to speed up the processing of reports. You can get more accurate results by increasing the sample size.
*Click the sampling icon and adjust this slider to the right to increase the sample size and thus the accuracy of your data.
- Upgrade to GA premium. Upgrading to GA premium comes with the perk of unsampled reports. If your site is sufficiently large that sampling is a common issue with reports, consider upgrading, but good luck getting the $150K yearly cost approved by your boss, I’ve never been able to do it.
- Use smaller time spans. Viewing your reports month by month reduces the chances of sampling becoming a problem because there’ll be a lot fewer sessions in the reports. For example, if you want to look at your data over two years, you can create 24 individual monthly reports that’ll most likely not be subject to sampling. You can then aggregate the data with a spreadsheet. This is time-consuming but more accurate.
- Focus on just one dimension. Combining dimensions can lead to reports being sampled vs. real data.
There are a number of other options to overcome sampling, some more complex than others. The previous three choices are the most straightforward options, but if you’d like to go more in-depth with this stuff, check out this excellent blog post on sampling by Lunametrics, or a paid tool that can do the job nicely like AnalyticsCanvas.
Can you track any additional items that you are presently not tracking?
This question is all about maximizing the efficiency of your self-audit so you can make better decisions about your current and future marketing campaigns.
Here are some ideas to get you started:
As discussed above in the goal section, events represent user interactions that occur outside of the traditional pageview. These are frequently used to track user interactions that are important, but may not represent the same endpoint as a goal.
Examples of good events to track are:
- Accounts Created
- User logins
- Social media shares
- External Link Clicks
- Add to Carts
- Add to Wishlists
- Primary Navigation Clicks
- Click to Call
- Form Submissions
- Mailto Link Usage
Site search can be configured to tracking internal website searches. This can be used to log how often users use the internal site search feature and what search terms they enter. This can prove invaluable for understanding how users view the website and what they are expecting to find. They internal search phrases can also be valuable for external marketing efforts such as creating keywords for search engine optimization and pay per click marketing.
Custom dimensions & metrics
Marketers frequently ignore the ability to include non-standard data in their website’s reports. Yes, Google Analytics tracks its own set of default metrics.
But this information is often of limited use for segmenting your customers properly. In a marketing context, demographics is kind of a big deal.
This kind of segmentation is a tad more helpful than just analyzing the behavior of an unknown population.
Let’s say you want to check out the number of females viewing your pages. You’ll probably have this information stored on some kind of CRM platform.
Implementing custom metrics allows you to combine a standard GA metric like pageviews with CRM data to get more precise information about the behavior of your site’s visitors.
The more information you have, the more finely tuned your future marketing campaigns will be. In other words, segmenting your market like this helps you kick more ass.
Another crucial example is segmenting users by email. You can use GA to track visitors who’ve been referred to your site by clicking on an email and see how they behave while browsing.
To set this up, you’ll need to create a new advanced segment with the medium set to email.
Email advanced segment
You’ll also need to tag all Google Analytics-tracked links in your emails with the appropriate tags. Let’s look at a simple sample URL first:
- &utm_source= welcome email
By tagging your links like this, you’ll be able to set up custom segments that track user behaviour from emails. This URL would track how users on example.com behave when they click the welcome email for a September newsletter campaign. Quite juicy information, don’t you agree?
On the flip side, Google is quite clear that tracking by email is against its TOS.
“You will not (and will not allow any third party to) use the Service to track, collect or upload any data that personally identifies an individual (such as a name, email address or billing information), or other data which can be reasonably linked to such information by Google”.
The issue here is one of cost v benefit. You are highly unlikely to be penalized by Google, but you do bear that risk if you track by email. On the other hand, the risk could be worth it for the invaluable information you receive about the performance of your email campaigns.
Is Demographic Tracking In Place?
In addition to providing internal data on the background of users who enter the website, demographics can prove extremely valuable for third party marketing. In fact, Analytics demographics lineup 1:1 with the same available within Google Adwords, when the latter is used. So, if certain demographics (such as age groups or gender) are performing well or poorly on the site, adjustments can be made to ad bidding to better target (or not target) these groups. Meanwhile, beyond age and gender, the affinity groupings and in-market segments can be used to help determine the types of sites & users who should be seeing the ads.
In most current organic traffic report, the vast majority of all organic keyword data in Google Analytics being listed as “not provided”. As such, one of the best reports for search engine optimization was rendered essentially useless. However, Google has compensated for this by creating a connection in Analytics for Google Search. This gives us access to keywords, impressions, clicks, and organic landing page performance. While highly sampled and unable to be connected to other Google Analytics dimensions or goal tracking, it still serves as the single best source of organic traffic data available from Google.
Custom attribution models.
An attribution model (in Google’s words) is:
“The rule, or set of rules, that determines how credit for sales and conversions is assigned to touchpoints in conversion paths.”
Unfortunately, the default attribution models in GA are way too simplified. Simplicity is always a nice thing to aspire to, but human beings are complex creatures, and their behavior on websites (and offline) is often complicated to understand and model effectively.
Consider the standard last non-direct click attribution model that Google Analytics uses in its reports. This model is way too simplified because it entirely ignores direct traffic conversions. Instead, all credit is given to referrals.
By its nature, the non-direct click model undervalues your marketing efforts and gives credit to everyone but you. Kind of harsh, right?
Not exactly a fair model of conversions.
The way around this is to customize the existing attribution models. This way, you can start to look at custom credit rules instead of just simply saying “X number of conversions arose from non-direct traffic” and assuming that this is an accurate statement.
Building an accurate custom model is extremely time-consuming, though. If you have the resources you can look into hiring an expert in this field to help. Or you could always sign-up for a premium GA account, which contains a useful and more accurate data-driven attribution model.
How data-driven attribution works. A much better approach than default models.
The key takeaway here is to recognize that there’s always opportunities to track more relevant information in Google Analytics for developing better marketing campaigns.
You should now have a clear idea about how to self-audit your Google Analytics. In summary, effective GA troubleshooting comes down to four key points:
- Make sure your GA tool is properly configured. This means setting smart goals and using filters intelligently.
- Ensure Google Analytics is set up across all pages on your site. In other words, make sure tracking code is omnipresent and structured correctly.
- Inform yourself of the common GA data issues that can cause you to make poor marketing decisions. Always question what you are seeing on your reports.
- Get the most out of your analytics by tracking non-standard metrics and utilizing custom attribution models.
Don’t forget to download your completely free Google Analytics Self-Audit Checklist below this article.
Do you have any of your own unique GA troubleshooting strategies? Has this article been useful to help you self-audit your analytics? Let me know in the comments section!
If you enjoyed this post, share it on social media.