Attribution Modeling with Google Analytics – 101
Connect With Us

Attribution Modeling with Google Analytics – 101

August 28th, 2013, By

Attribution modeling has become a hot topic in marketing analytics. In this post, we’ll explain the basic concepts, briefly touch on the latest attribution modeling methods, and discuss how to use them effectively to optimize your marketing efforts.


The Basics

Attribution means determining the source, or sources of your target outcome. In the context of marketing analytics, this means accrediting conversions to the marketing channel that drove them. I.e. if someone clicked on a banner ad to reach my site, and made an online purchase, then the sale should be attributed to my display marketing campaign.




Let’s take the following scenario for example:

  • A visitor clicked on a link to your website via one of your Facebook posts that was shared by one of her friends.
  • Later that week, she saw a display ad via a Remarketing campaign, followed it to a landing page, and learnt more about the offer.
  • Finally, she returned to the website directly (by typing in the address), and made a purchase.

Attribution Modeling is the act of deciding which marketing channel should be given the credit for the conversion, in order to assess the cost / benefit of each channel.


There are several possible approaches. To name a few:

  • Last click – credit should be given to the click that directly preceded the conversion – in this case, the direct channel.
  • First click, which ignited the first interaction and awareness of the brand, should be given all the credit – in this case Facebook (social).
  • Last non-direct click – assuming that the visitor came directly to convert after already being won by the previous engagement – i.e. via the Display ad.
  • The Linear model – credit should be split evenly between all the touchpoints along the way.
  • First and last click – split the credit between first and last click
  • Many variation on those ideas

The Evolution of Attribution with Google Analytics

By default, GA attributes conversions to last non-direct link.

Two years ago, GA came out with Multi-Channel-Funnels – a set of reports under the Conversions section which  shows the full paths of traffic sources that led to conversions (in the example above – Social -> Display -> Direct -> Conversion.


The latest feature from Google, attribution modeling, lets you apply different models, or rules, which determine the allocation of goal credit between the different channels that made up the full path to conversion. You can choose from a set of preset models, or create your own attribution rules. Finally, you can compare different models alongside each other.




Why should I use Attribution Modeling?

Using Attribution modeling lets you assess the value of all your marketing efforts, taking into account the entire succession of site engagements that led to a conversion. This, in turn, allows you to make smarter decisions in allocating your budget. By contrast, relying on a simplistic method such as last click, may lead you to take a narrow and incomplete view on your marketing channels.


Which model should I use?

Choose (or create) your attribution model according to your business model and assumptions. For instance, if your typical transaction is a quick decision which doesn’t require a long thought process, then the Last Click model makes sense. If you’re a new brand and your main efforts go to creating initial awareness, then you might want to accredit the First Click more than the others.


Why is comparing different models so beneficial?

Looking at several attribution models together can give you highly valuable insights.

For instance, let’s say that using Last Click, as your primary model, you find that social-campaign traffic performs weekly. However, when comparing with the First Click model, you find that social campaigns are the main source of bringing visitors, which ultimately convert, to your site for the first time. What can you learn from this?


First of all, you’d realize that social media is definitely a place to invest. Furthermore, you’d realize how to use this channel effectively – since visitors coming from social may not be ready to convert, you might consider focusing on making sure they return to your website, e.g. offering them to subscribe to your newsletter or social media channel, rather than going with a hard-sale approach.


Attribution Modeling enables better, more informed, optimization of your marketing investments, as it gives you a fuller view of the ecosystem.