April 16th 2020

Why do you need a Google Analytics Audit?

Every successful SEO campaign begins with data. Data helps you identify the issues, make informed decisions on what approach to take in resolving these issues, determine priority and most important; measure if your SEO efforts have been successful. 

As an SEO agency, every client campaign we onboard begins with a performance benchmark. The purpose of this benchmark is to evaluate how close the client is to achieving their goals so we can determine the success or return on investment (ROI) after our recommendations have been implemented. If the client’s Google Analytics (GA) account has been incorrectly set up then it becomes very difficult to prove any value with flawed data.

Clients often approach us after they have created a SEO forecast based on data from GA. In this forecast, the client has projected what level of growth is expected from organic traffic and set targets based on these figures. They expect us not only to validate these figures but also work with them on a strategy to achieve it. This is another reason we recommend auditing your analytics account to ensure your traffic projection and targets come from a clean, accurate data source especially if you are making financial decisions from these which most companies do. To give some context to the margin of error we see; 80% of the analytics accounts we audit are incorrectly set up.

80% of the analytics accounts we audit are incorrectly set up

Even if you are not working towards validating a forecast or target, you still need clean data to make the right decisions. Google Analytics is a great tool for understanding how people use your website. However, like most tools, the data you pull out from it is only as good as the data that it records.  The standard out of the box set up isn’t the most efficient and it usually requires some customisation to get the best out of it. That is why we have created a free GA Audit template which will enable you to perform a comprehensive data hygiene exercise. Our template highlights 30 issues that might be holding your account back and how to fix them.

Get a copy of your Free Google Analytics Template Here

Our audit template is formatted in a handy google sheet, to use this template just create a copy.

Most Common Errors in GA

Do you trust your data? These are some of the most common errors we see with accounts we audit. If any of these apply to you then you need to conduct a data hygiene exercise on your account. These are all included in the template.

1. No filters

Filters are used to limit or modify data in a reporting view. This helps make your data more precise and concise. The uses of filters can be categorised into three main benefits;

  • Segmenting data (for a more granular view). Google Analytics (GA) only shows you a sample (subset) of your data (except you have Google 360 which gives you access to bigger sample size). The more dimensions you add, the smaller the sample size.  If the sample size is too small it can lead to decisions being made on inaccurate data. Filters are a great way to segment your data without compromising on your sample size. You can find out more GA sample limits here
  • Custom views. Depending on the needs of your business you might want to create custom views either to limit the information you share with people or be able to access granular or specific information quickly. An example is creating specific views for traffic from different countries or subdomain
  • Cleaning up data. You can use filters for data hygiene a good example of this is excluding internal IP addresses from being recorded as organic visits. Other common examples include removing query parameters from URLs, changing campaign URLs from uppercase to lowercase.

Filtering your data permanently alters that data set so ensure you always have an unfiltered view of your data to avoid permanently losing important data.

“Always have an unfiltered view of your data and make copies there from which you can filter and customise”

2. Single view

GA automatically created one unfiltered view for every property in your account. However, you need at least three views. A master view which is usually a customised view with all your filters and customised report. A Raw view which is all your data unfiltered and untouched this is your back up view. Incase anything of any data lose you can create a new data set from your raw view and finally a test view. A test view is important for testing out any changes or modification you want to make to your data to ensure it works accurately before you roll it out to the master view this is another precaution to prevent data loss.

3. Wrong domain name

We come across this on accounts where the domain has been migrated from HTTP to HTTPS, however, the default domain name has not been updated in GA. This can lead to a spike in referral or direct traffic as Google with double counts traffic from the same domain or lists it as a referrer depending on the account set up. To correct this simply change your default domain name to the HTTPS version.

4. GA code in the wrong part of source code

Your GA tracking code should be placed on every page in the <head> tag. This is to ensure that when a user navigates to your site, the code is triggered as soon as possible. If the code is not triggered as soon as a user navigates to the page it can lead to the visit not being recorded. Bounce rates will also be incorrectly recorded as users will be on the page for sometime before the tracking code is triggered.

5. Incorrect goals & events set up

Goals and events are essential in determining the success of any campaign. If these are not correctly set up then it will either be difficult to analyse if KPIs (key performance index) are being met or decisions will be made on the wrong data.

6. Referral exclusion list

If you have a sub-domain and have implemented cross-domain tracking or send users to 3rd party sites (usually payment gateways), you need a referral exclusion list to stop that traffic from getting double-counted. Creating a referral exclusion list is easy. In the account section of your account navigate to property > tracking info > referral exclusion list>add referral exclusion

Example of referral exclusion list in Google Analytics

7. Enhanced ecommerce

Some clients wrongly assume that setting ticking the enhanced eCommerce button in GA means that it has been set up correctly and all checkout data will be correctly recorded.

example of enhanced ecommerce wrong set up

To accurately set up enhanced eCommerce you need to set up a Datalayer to collect the checkout data and send it to GA. This information will populate the check out funnel.

Example of an enhanced eCommerce data spec

8. Personally identifiable information (PII).

PII is any information that can be used to identify a specific individual. Some examples are email addresses, house address or phone numbers. To protect your user’s privacy, GA states in its policy that PII must not be passed on to Google. Failure to comply will result in your account being closed. You are also breaking GDPR rules by sending PII to Google. If you find PII in your data you should work with your dev team to obfuscate the data.

9. Bot & Spam traffic

A bot is an automated script (non-human robots and spiders)  that is programmed to carry out tasks on the internet. Not all bots are harmful however they are not real users and so visits from bots can inflate and skew your data. Bot activity can be found in your referral traffic report. To identify bot traffic, look for unusual activity and patterns (one of the patterns to look for is an unusually high bounce rate when there is a spike in traffic). Harmful bots (i.e content scrapers) should be blocked from access the site directly from the server. Non-harmful bots can be filtered out in GA. To learn more about bots read our what is a bot article. We also have a handy step by step guide on how to filter out bot traffic in GA.

Spam traffic is very similar to bot traffic because it has a similar effect on data. It creates an unusual spike in visits and high bounce rates for the duration of the spike.  Spam traffic can also be excluded using filters or segments to exclude the sessions.

Example of unusual data spike in Google Analytics
Example of a high bounce rate in Google Analytics
Example of how to filter out data in Google Analytics

10. Content Grouping

Technically not an error but an enhancement, content grouping enables you to sort your site’s content into a logical structure that reflects how you think about it. You can group your content by either page type or page use. Setting this segmentation will give an in-depth view of how sections of the site are performing and make future diagnosis easier.

11. 404 tracking

There are many ways of tracking broken pages on your site. You can monitor broken page on Google Search Console or perform a site crawl to determine how many 404 pages are on your site. However, what is even more useful is discovering how many times your users see a 404 error and how they interact with your site following that. This will enable you to make an informed decision on what to do with that page. Internal 404 tracking can be set up with custom reports.

If you’d like our assistance with auditing your Google Analytics accounts, feel free to reach out to us.