8 simple fixes to clean up your analytics data

Data hygiene is vital, but most companies are still not getting the basics right. When was the last time you did a little spring cleaning yourself?

Here are eight simple things you can implement or tweak to get your data in better shape.

1. Filter out irrelevant traffic

Filters in Google Analytics are underused. You should be filtering out internal traffic (especially if you have a big call centre, for example), as well as spam traffic and bot traffic, to get a more accurate picture of how your customers are using your site. And don't forget to exclude any staging sites that might be using the same GA tracking code.

2. Track your conversions (and micro-conversions)

Conversions and 'micro-conversions' are the bread and butter of your Google Analytics account.
Micro-conversions are the steps along the journey to a conversion (so, for instance, adding items to a basket or wishlist, or the steps leading up to submitting a form for a lead gen site), and you need to be tracking them so you can analyse the journey your customers are taking to determine and eradicate any 'blockers' preventing them from converting. Here's some guidance on how to sent up Event Tracking through Google Tag Manger, which can help you track those micro-level actions.

3. Use the optimum number of GA 'views'

There are two common problems that occur here; either you have the bare minimum number of views (1), or you have too many.

The more views you create, the greater the risk of data sampling (when using the free version of Google Analytics at least). The majority of the time, an excessive number of views are created when the same outcome can be achieved through effective data visualisation or segmentation.

We recommend having three views:

A raw view

This is an unfiltered view with no configuration applied; this will act as a back-up should any configuration changes inadvertently stop tracking.

A test view

This will be used to test any configuration changes to ensure that said changes have had the desired effect, and also to provide a preview of the data before the configuration changes are promoted into the main view.

A main view

This view will be your main reporting view with your full configuration applied. This should be used for analysing, gathering and disseminating data.

4. Make use of annotations

Annotations are markers you can add into the data to pinpoint when significant events happen that might affect traffic. For instance, you could annotate when a site update was released, or when a marketing campaign went live. These are really valuable, not only for your analyst to see the value of this activity and to pinpoint any variances caused by these events, but also for anyone new to your Analytics account to see the impact of any external factors that wouldn't otherwise be obvious.

5. Reclassify your events as interactive or non-interactive

An interactive event is when a user takes an action, such as clicking a button, or moves to another page. Whereas a non-interactive event would typically be used to track page engagement so for instance when someone has scrolled below the fold, has spent time looking at products or engaged with an embedded video.

Making the distinction will allow you to include the non-interactive events as part of your bounce rate calculation making it a much more valuable metric.

6. Track calls from your site

There are a number of tools on the market you can purchase to track if a call is made from your site. These are invaluable for lead generation businesses in particular, as a large number of users will look to call you rather than using your form. If you don't have call tracking software on your site there is no way of attributing these conversions back to any marketing efforts. And if they're not implemented correctly, it can lead to a landing page of (not set) and a 0% bounce rate for these pages.

7. Ensure payment gateways are being properly handled

If you use any sort of third party payment gateways such as World Pay or Paypal, you need to ensure the traffic is handled correctly in GA. If not, when a user goes from your website to paypal or other payment gateway to make the transaction the revenue will be tracked as referral traffic from Paypal, essentially making your previous marketing tracking redundant.

To do handle the data properly you can block that payment gateway as a referrer, and when you send them to the payment gateway you can set up a virtual page view that you can track through GA. This will allow you to virtually include the payment gateway as a funnel.

8. Check your settings are correct

This is a really basic one, but it's important to make sure the correct time zone, currency etc. are set up in your Analytics account. Here's some guidance on how to change your Google Analytics settings.

If you put these 8 tips into practice, you'll be in a stronger position to start getting better insights from your data, and prioritising the right tests and improvements moving forward.

Whilst you're here, why not learn why the web metrics you're measuring are probably wrong.

How to implement personalisation using machine learning