Google Analytics account setup – 3 recommendations

Within the Strategy & Insights team, we always have a key focus on providing data-driven recommendations to help align digital marketing strategies. Whether we’re working on a Discovery research project or undertaking a CRO campaign for a client, there is always a key focus on Google Analytics and how we can use data to drive our decision making.  

Why Use Google Analytics?

With this considered, we use Google Analytics on an almost daily basis for all clients, in order to track how their websites are performing. As such, it’s important for us to ensure that the data shown in the different reports are as accurate as possible.

If a Google Analytics account hasn’t been set up correctly, it can lead to a number of different scenarios, which can cloud your data, and in turn, challenge your decision-making process. Examples of this leading to incorrect data and reports include:

  • Inaccurate traffic levels
  • Incorrect eCommerce and revenue figures
  • Traffic sources or marketing channels being pulled through as the wrong source
  • Spam or bot traffic skewing referral reports
  • Incorrect conversion rate reporting

Whilst we’d suggest getting in touch if you’re really concerned about the data tracking on your website, here are three quick, handy tips for improving the quality of data you’re monitoring for your website.

1 – Views

As part of their extensive Google Analytics training courses, Google recommends that a website should have at least 3 views on their account: unfiltered data view, test view, master view.

  • The master view acts as the view you would use to carry out your reporting, using the required filters and goals set up, amongst other attributes.  
  • The test view can be used to test changes to different attributes, ahead of then using them in the reporting master view.
  • The unfiltered data is as it sounds, a view not containing any filters, with a very basic account setup, so that all historical website data can be viewed there as a backup and reference points.

We have found across many clients’ accounts upon our first examination that they may only have one view in their Analytics account, meaning that they don’t essentially have a ‘backup’ of their data and could permanently impact their reporting if they were to add filters or make ay account changes.  

However, on the other hand, we often find some accounts where there are more than 10 views, which can lead to a confused experience working to establish which is the main reporting view and can often lead to members of the client’s business on different pages, as everybody is looking at different data sets!

As per our previous point here, we would recommend in most circumstances having these 3 views and potentially renaming or removing other views which aren’t used for your main reporting. It’s a simple change to make, but it’s a quick job to ensure the health, accuracy, and safety of the website data you manage.

2 – Filters

Filters are an important part of your Google Analytics setup, allowing you to filter out any traffic which could ultimately be skewing your reports or to manipulate how data is pulled through.

The range of filters available varies widely, from excluding traffic from certain IP addresses to ensuring that the full URL and domain is pulled through to content reports.

As a bare minimum we would recommend using the following filters to exclude traffic and prevent more noticeable skewed reports:

  • Exclude traffic from your company offices

It is likely that internal traffic testing a website or customer service staff talking a customer through pages over the phone would spend longer on the website.

  • Exclude traffic from your digital agency

As a Digital agency, we are likely to spend a good period of time on our clients’ websites, whether that be testing, carrying out a UX and design audit or a page speed test.

  • Exclude potential bot spam traffic

If not filtered out, this spam traffic can pull through as a referral but have engagement levels showing 100% bounce rates with sessions lasting 0 seconds, which obviously wouldn’t be possible for a user to do.

  • Force URLs to display as lowercase

Its human nature for typing errors to throw in the odd uppercase letter into a URL. Without this filter, it is possible that duplicate content pages could be pulled through into your content reports. For example, www.clicky.co.uk/Contact/ would pull through as a separate content page to www.clicky.co.uk/contact/ otherwise.

Adding these filters to your Google Analytics account and giving them a good test might take you less than an hour, but it’s time well spent. The alternative is that your digital marketing team might be spending a whole lot of time optimising your website towards ‘users’ who are actually your own internal staff, visiting the website from work.

3 – Channel groupings

By channel groupings, we are, of course, referring to the different default sources or ‘channels’  which send traffic through to the website. These default sources (or as Google Analytics defines them, channel groupings) are as follows:

  • Direct
  • Organic
  • Social
  • Email
  • Affiliates
  • Referral
  • Paid Search
  • Other
  • Display

Within our health check and then as part of our further Acquisition reports, we often come across and make the following recommendations to create new or modify existing channel groupings.

With paid traffic in particular considered, it is likely that your audiences have different researching and buying behaviours – whether they are searching for more generic or brand specific terms – so splitting these visitors out will provide more accurate reporting on this audience’s behaviour and how they engage with your website.

  • Paid search → you can split out this traffic to create 2 new channel groupings: ‘Branded paid search’ and ‘Generic paid search’. This will help you keep an eye on how users behave, whether they came from a competitive term, or a branded company keyword search. It’ll also help you see how much branded PPC traffic your business is paying for too.
  • Organic search → you can update the existing channel grouping to include traffic from other search engines, which are normally missed by Google. Such search engines include Yahoo, DuckDuckGo and Ecosia are often not included in this category, but certainly should be!
  • Other → this is essentially Google’s ‘uncategorised’ bin. This is where you’ll find all traffic sources which Google is a bit confused about. Normally, this traffic will belong somewhere else, keeping on top of this can help to keep a clean, tidy data set. Common offenders you’ll find in this list are Instagram, Pinterest, and some popular referral websites.

These findings are a very top level summary of common areas of improvements we identify and recommend to our clients straight away. If you’d like to investigate your website data and understand how users are interacting with your website, and where they are potentially dropping off, get in touch with our team today on 0800 222 9300 or email hello@clicky.co.uk.

Written by Rhodri Lloyd

Digital Strategy Executive

Follow Rhodri on Twitter