Typical Issues When Working with Google Analytics: How to Fix Them

Sophisticated internet marketing is impossible without analytics. But what do you do if your statistics are displaying incorrectly?

Below, we’ll take a look at the technical difficulties that you can experience while working with Google Analytics. These range from interface limitations, to tracking configuration errors, to receiving inaccurate data.

Incorrect Display of Parameters and Numbers in Reporting

Referral Traffic from Inappropriate Sources

From day to day in Google Analytics reports, you may find extra traffic from unknown sources. For example:

Subdomain Referrals: in the case of several subdomains,  it’s possible to switch between them. The stats will then show this: sub.domain.com / referral.

Payment System Referrals: if you pay for orders on your site through a third-party system, in the reports you’ll see traffic related to … / referral.

Accounts.google.com Referrals: if your website provides a user signing in through your Google account, then after they log in the session is overwritten.

Such sources can create a great number of conversions. As a result, the real source that led this traffic is evaluated on false data. To solve this problem, the exclusion of transition sources will help. Include this item in your checklist for setting up your analytics to prevent making decisions based on false data.

To do this, go to the admin settings at the share level. Select the “Tracking” section and click on the list of excluded migration sources.

 

Add each domain individually that you don’t see as a source of migration in your reports:

 

 

Note that this setting only works with analytics.js and gtag.js codes.

If you have ga.js installed, the setting should be made itself.

Once this setting is done, it will not be terminated and the merited conversion will be linked to the correct source.

Markings in Google Analytics Reports

Briefly about the problem: Google Analytics can sometimes hide data because of privacy. This is usually the case when “(not provided),” or “(not set)” appears, instead of a target’s age or location, for example. It’s often encountered by SEO professionals when analyzing search queries. Some reasons you might have this include:

1. Lack of connection between Google Advertising and Google Analytics or an error in its setting.

2. Problems with UTM tags:

• Unplugged auto-marking in Google Advertising
• Mistakes in syntax were made at manual markup (parameters themselves are written with errors)
• Mixed tag is used incorrectly

3. Features of the site – for example, the parameters are converted to lowercase, 404 not found, or the tag is cut off – because these nuances will be incorrectly transmitted (or completely lost) on the parameter “gclid” or hand tags.

4. Google may not receive enough data about search systems. Because of this, said traffic stays unidentified or written to the referral.

5. Finishing the session at the end of the day – this is the default and is followed by overwriting the session. Here, there may be changes in the parameters.

Sampling

To put it simply, sampling is a manipulation of the source data used to identify a smaller set of data similar in properties to the source data.

What’s really going on: if you have a large number of statistics, Google Analytics takes only part of the data to multiply it several times by  100%.

Here’s how you can avoid this, from simple steps to an advanced level, with maximum convenience:

• Use smaller time periods in reports: build a report in the section of each month instead of an entire year and analyze separately
• Create separate resources for different sources (but such data will be difficult to combine)
• Upload data using add-ons such as Google Analytics. Replication will also take place, but it will be easier to upload data in small periods using sheets

Problems with Advertising Systems

Connection to Google Advertising

There may be errors in the settings of Google Analytics and Google Advertising, so serious data discrepancies may occur. Incorrect statistics will lead you to the wrong conclusions and ineffective decisions.

What mistakes may occur and how to correct them:

1. There is no connection between Google Advertising and Google Analytics

In Google Analytics, go to “Administrator” – “Resource” – “Connection with Google Advertising”:

Select the desired advertising account:

You can also connect through the Google Advertising interface: “Settings” – “Linked Accounts” – “Google Analytics.”

2. Google Advertising auto-tag is disabled

You can enable it in the “Account Settings” section by checking the “Automatically Tagging” box.

3. Incorrect use of several markings

If automatic tagging is not enough for you, check the corresponding item in the “Property Settings” section of Google Analytics:

This will allow you to manually set some UTM tags.

But make sure you fill in the labels correctly. For example, if you type a value for a campaign that is different from Google Advertising, your data will be separated in Analytics.

Parsing Expenditures of other Advertising Systems

Once you have set up a connection between Google Analytics and your Google Advertising account, the reports will show data that precedes your visit to the website. It cannot be gathered by the counter – for example, the number of clicks and the cost of clicking.

But when using other advertising systems, it’s necessary to obtain similar data from them. For this purpose, you can set up data import using additional services.

Visualization

Anyone working with Google Analytics knows how important and useful data visualization is. It makes information easier to understand and often reveals problem areas that may be invisible to the naked eye.

Everyone who has worked with Google Analytics has made sure that the possibilities for building charts are very limited. To solve this problem, you may install external tools.

You need to choose a tool based on your needs. Keep in mind that when using paid tools, their costs must be accounted for in ROMI. This means that for smaller projects, additional tools are often not allowed because they cannot pay for themselves.

Problems with the Quality of Targets (Marketing Investment Return)

Duplication of Transactions

Wonder why your transaction data in Google Analytics is being multiplied? It could be because you and/or your visitors refreshed the “Thank you for your order” page. Moreover, the problem may be in the implementation of e-commerce.

To correct it, determine what exactly is the reason it’s happening, then:

• Make a configuration with GTM
• Change the logic of the “Thank you for ordering” page to dodge updating it
• Do not link up the “Purchase” code of e-commerce to the website and upload transaction data using Measurement Protocol

Data transfer from CRM (Measurement Protocol)

In fact, this is a solution. The issue, however, is that Google Analytics doesn’t have any data about the actions in the future of your funnel.

This forms some limitations and errors in analysis:

• Lack of data on transactions made off-site (by phone, offline, etc)
• Inaccurate transaction data due to lack of data on cancelled transactions
• Lack of LTV (Life Time Value) data, etc.

The solution: building a private analytics system using one of the above-mentioned methods.

You may use off-the-shelf solutions, or you can build your own analytics. This depends on your resources:

• Use off-the-shelf solutions, if your budget allows
• If you have a full-time developer, assign them the task of implementing post-requests via Measurement Protocol – this option may be both convenient and profitable for you

Conclusions

When using Google Analytics, anyone can encounter a number of questions or issues. Yet each problem comes with a solution:

• To avoid bugs in tracking configuration, read the FAQs and Help carefully, and prepare clear terms of reference for developers
• Exporting data to additional storage and analytics systems to help avoid sampling
• Get ahead: regularly check your site for redirections to save tags in links and duplicate transactions
• The limitations of Google Analytics aren’t that bad and there’s a solution to every task
• Build your own analytics system and read ROMI