google analytics attribution modeling graphic

What is Attribution Modeling in Google Analytics?

Attribution Modeling in Google Analytics assigns values of credit to defined channels within a buyer’s journey to conversion. That’s a lot to take in – so think of it this way: Digital marketing experts and data analysts attribute credit to the parts of the journey they believe won the conversion into a business lead or customer using attribution in Google Analytics.

Can I Use It If I’m Not a Data Analyst?

Absolutely – Google Analytics does most of the hard work for us. It collects massive amounts of data and filters it into handy reports through which we gain insight into the behavior of visitors to our site. If we want to set up attributions to discover more about what our customers are up to on our site and at what point they decide to become our customer, we need only to dive into the Conversions menu and explore Multi-Channel Funnel Reports and Attributions.

Among the wealth of features available in the Conversions menu are reports that show us the buyer’s journey in different ways. Google Analytics has a predefined set of attribution models available through which we can filter the data in those reports even further. There is no set up required to use the default attribution models provided by Google Analytics – but for those who want the deepest knowledge of their most powerful marketing channels, some assembly is required.

Concepts You Must Know to Understand Attributions

Though you may already be familiar with some of the key concepts behind defining attributions, fully understanding the complex nature of such a customizable tool requires confidence in the key concepts of Google Analytics data analysis. Be familiar and comfortable with the following terms before moving forward.


Conversions are the actions in a marketing funnel that convert a prospective customer into a real customer. There is no one specific action that is rigidly defined as a conversion, but some of the most common actions include filling out a contact form or sending an email to the sales team.

contact form for copywriting showing conversion

Once the customer hits Send, he or she has converted.

Digital marketers often set up goals in their data analysis tools around conversions. While conversion rate is only one among many essential factors of marketing success, turning an interested party into a potential customer or client is incredibly valuable and easily measurable. But since there’s also no way of defining what makes a potential client a potential client, we have to set some kind of bounds in which we can identify site visitors as potential clients. We do so with goals.

In the world of marketing, goals are user-defined target actions for the potential buyer to undergo. Goals are infinitely customizable – they can be something as simple as a social share to something as complex as a page visit from a search using a specific keyword following a link.

The main functions of Google Analytics outside of the Conversions menu define conversions as goal completions. Inside the Conversions menu, Google Analytics automatically adds e-commerce transactions AND goal completions as conversions. Knowing this distinction will help you create more concrete conclusions from the data you collect on conversions.

Multi-Channel Funnel Reports

Multi-channel funnel reports show you through which channels your customers move into conversion. The reports define the path, from beginning to end, that individual converted site visitors have taken up to the point of conversion – and show you every step along the way.

When you click on Top Conversion Paths in the Multi-Channel Funnel Reports menu, you’ll see several rows like this:

conversion path in a multi channel funnel report

This is a long conversion path, but it shows us exactly what the customer did to reach conversion. The [x #] means that particular action happened twice or more in a row. We can assume this customer found the site via a paid search ad, came back directly later, looked again and found the site twice more via paid search, came back twice to learn more directly, reached the paid ad again, found the site through a link, then came back directly a final time to convert.

The report shows us quite a bit about the customer’s path, but the information is superficial. We don’t know what the customer was thinking through each one of these channels, nor do we know what exactly spurred a conversion. That’s where attributions begin to come in handy.


Attributions are the credits we give to different actions taken by visitors to our site. Remember, attributions are different from conversions and multi-channel funnel reports. We assign attribution to find out which specific pieces of the buyer’s journey are the most effective and how we can maximize that effectiveness.

examples of channels through which attributions are made

The channels through which attributions are made.

When we assign attribution, statistics are shifted to reflect the credit we’ve given to certain channels and interactions so we can get a clearer picture of the impact of spending money on digital marketing. You may discover through attribution that your CPA (cost per acquisition) is less than you may have previously thought. This occurs because other interactions that may cost you money are actually happening after the action to which you’ve attributed credit for a conversion. That’s how attributions help us look deeper into our marketing.

Understanding the Significance of the Default Attribution Models

An attribution model is different from an attribution. The attribution is the credit we give to actions that lead to conversion. Attribution modeling is how those attributions are distributed according to the rules we create.

You’ll find the Attribution Model Comparison Tool under the Conversions menu after clicking Attribution. In the Model Comparison Tool, we can begin to make attributions to the different parts of the buyer’s journey. We can also compare one attribution model to another to draw more profound conclusions.

The free version of Google Analytics has a limited set of default attribution models that will help you begin building a picture of your most powerful converting tools. They include the following model types.

Last Interaction

last interaction attribution model icon

Assigns credit for a conversion to the very last action taken by a converted site visitor. Use this model if the data you’ve collected from customers signals that their decisions were made based on the last piece of content with which they interacted. You can then set the cost of your conversions to reflect the entire course to conversion.

Last Non-Direct Click

last non direct click attribution model icon

Hands credit for a conversion to the action BEFORE the actual customer conversion. Take advantage of this model if your data shows that customers made the decision to get more information or purchase based on content just before their final decision. This assumes that your customer already decided to make a purchase before directly going to the site to convert.

Last AdWords Click

last adwords click attribution model icon

Assigns credit for a conversion to the last paid search before the customer converted. Employ this model to focus on the efficacy of paid search and compare which of your paid search campaigns have the best cost to return on investment ratio.

First Interaction

first interaction attribution model icon

Gives conversion credit to the very first interaction of a converted customer with your website. Switch to this model to see how much it costs to grab customers that will eventually convert. You can also monitor which channels do so and if you can take advantage of untapped channels to save money on initial interactions.


linear attribution model icon

Distributes credit for conversions across each channel that assisted in a conversion. This model will help you determine which channels fall into the funnel in similar places and which ones get the most attention. It will also give you the most unbiased picture of your CPA.

Time Decay

time decay attribution model icon

The time decay model is special – it operates under the assumption that each day that passes after a customer’s initial visit to a site, the chance of a return visit splits in half. Attribution credit shrinks to lower percentages based on the time decay, which gives less credit to early interactions and more credit to later interactions the longer the customer’s path becomes.

Position Based

position based attribution model icon

Offers 40% credit to both the first and last interactions, then pushes the remaining 20% credit evenly across the remaining channels in the conversion path. Use this model if you want to take a closer look at how the majority of users are introduced to your site and what pushes them toward their final decision. Your CPA will reflect a more complete picture, but will do so with more bias than the linear model.

These models are useful for very basic analysis of cost for parts of the conversion path that have a value that’s tough to track.

Creating Your Own Attribution Models

All businesses are unique, and so is the odyssey that the buyer takes toward making a decision about the business’s product or service. Business goals may not align with competitors, and marketing approaches may differ drastically. In all of these cases, custom attribution models are useful for honing in on how customers react to important parts of your website.

Understanding the Power of the Attribution Model Creation Tool

The attribution model creation tool is massive. You can select from lots of different criteria to define on a very detailed level how your customers interact with your site and where they become inspired to convert.

custom attribution model creation tool

A custom model set to record 1.5 times the credit for initial visits to the site from Social channels

Access this tool by clicking the gray, italic Select Model option after opening the Attribution menu and clicking Model Comparison Tool. At the bottom of the Select Model menu, start creating by clicking Create new custom model.

You can assign an outrageous number of rules to your desired specificity by using the AND and OR statements that make up the bulk of your customization. Because so many metrics are available to feed and filter your data, it’s easy to see how quickly the scope of this tool becomes astronomical. You’ll have to stay grounded in the big picture while examining the finer details of attribution.

Example Scenarios

To keep from getting too lost in the minutia, try to return as often as possible to broad goals for your business when making new attribution models. Take some of these situations for example:

  • Business with heavy organic search traffic may like to attribute credit to specific keywords by:
    • Using the First Interaction Model as a base
    • Including Source specified as Organic Search as a rule
    • Specifying the targeted keyword(s) that originated that search

custom attribution model rules for heavy organic traffic


  • Business focusing on social traffic that leads to conversions may attribute credit to social channels heavily by:
    • Starting with the Position-Based model
    • Adding Social as a condition
    • Defining where in the path the credit should go

custom attribution model for first social interaction


While these scenarios may be unreasonable for some businesses, they provide a loose concept of what is possible with Google Analytics attribution models.

Creating Attributions That Matter

To really dig deep into where you should apply your credit, you’ll have to get on a personal level with your audience. There’s a large time and effort investment in finding out how people interact with your site below the surface, but once you’ve collected the information you need and your custom attributions are set, you’ll have the most complete Google Analytics report among your competitors.

Ways to Collect Information

Get to that hard to reach information about your audience by:

  • Including a simple survey question or two in your contact form
  • Ask visitors some follow-up questions on your Thank You page
  • Politely request answers to a survey in a follow-up email

Find some way to ask your customers how they heard about the site or what caused them to want to become a customer. Ask them if they would recommend to a friend and why. From this data, you can start to assign more credit where the power of your investments is actually going and watch how those specific channels contribute to the overall buyer journey.

Also, watch how others are using Attributions and use or tweak their methods. Follow creative ideas like this mathematical approach to determining the strength of visits. There’s no end to the possibilities of the Google Analytics Attribution Modeling comparison tool and custom model creation tool – you need only to learn how to harness the incredible amount of data that waits locked away in the chambers of Google’s servers.

Know who’s skilled at using Attributions in Google Analytics? The digital marketing team at Leverage Marketing. Talk to us if you need real marketing results with the data and expertise to back it up – or empower yourself with the solid inbound marketing knowledge we share by signing up for our newsletter below!

2 replies
  1. Laura Wyss
    Laura Wyss says:

    Nice Post. Attribution modelling is a complicated affair. It’s so complicated that most e-commerce stores don’t even dive into it. The hidden part of the story is that if they do, they can reap great benefits out of the model as it is highly effective for growth prospects.

    • Eric Ysasi
      Eric Ysasi says:

      Hi Laura,

      When you’re right, you’re right. When you can attribute your sales accurately to your traffic channels and campaigns, you can start focusing more time and money into what works and stop funneling resources into what doesn’t. Plus, you only have to set up attribution when you have a new channel to monitor – once it’s done, it’s done!

Trackbacks & Pingbacks

    Leave a Reply

    Want to join the discussion?
    Feel free to contribute!

    Leave a Reply