Model Comparison Tool Google Analytics

iStock_000061152844_650x250

There's a key question that most marketing teams want to know the answer to:

"How should we be attributing credit to all our different marketing channels?"

One answer that provides helpful insight into this issue is Google Analytics' Model Comparison Tool.

What Is the Model Comparison Tool?

As its name suggests, the model comparison tool allows you to look at all your channels and compare three different attribution models with one another in Google Analytics. This helps marketers determine how a marketing channel can be valued from different perspectives.

For example, you could compare how your paid Facebook advertising stacks up when you apply Last Click attribution with Linear and Time Decay attribution.

model comparison shot

You can use this tool to:

1. Use pre-set attribution models, including:

Last Click – This model looks at the last channel the customer used to arrive at your site, including a direct hit (when the user types your URL directly into Google). The last click gets all the credit for the conversion.
Last Non-Direct Click – Similar to "Last Click," however this model discounts a direct hit as a touchpoint if it is the last touchpoint before a conversion.
Note: this is GA's default attribution model which it uses in all the other reports it offers.
Last AdWords Click – In this model, the last AdWords campaign interaction gets all the credit.
First Click – In this model, the first channel that got the customer to your website within whatever lookback window you chose gets all the credit.
Note that if you've set your lookback window to 30 days, any channels that may have gotten a customer to your site 45 days ago will not be shown – thus, pick a lookback window that makes sense for your business.
Linear – In this model, each channel that brought a customer to your site (taking into account clicks or impressions or both) before final conversion gets equal credit (again, within your chosen lookback window).
Time Decay – In this model, each touchpoint gets an increasing amount of credit as the touchpoints get closer to the moment when the conversion happened. This model is somewhat similar to Last Click, however it takes into account other channels which may have contributed to a degree.
U Shape (position-based) – In this model, the first and last touchpoints within your lookback window tend to get the most credit, as these were the "introducing" and "closing" touchpoints – i.e., what familiarized the customer with your product and what caused them to convert. Any middle touchpoints either get a portion of credit each or no credit. Thus, if the percentages of credit were graphed from the first to last, it would result in a U-Shaped graph. U-Shape attribution can emphasize whatever position you choose. It is a base model that you can select when creating a custom model, allowing you to assign whatever percentage of credit you want in the beginning, middle, or end of the series of touchpoints.

2. Create a custom model

In addition to the above default models, you can use the Model Comparison Tool to create, save, and apply a custom model that uses the rules you specify. This allows you to tailor models specifically to the set of assumptions you wish to evaluate in your conversion path data.

When you want to see which channels are benefiting from different types of attribution models without having to first change your whole process and existing reporting methods, the Model Comparison Tool is very useful. The key is to look at the movement, i.e. which channels are benefiting when you change your model, which ones are undervalued, etc.

As you explore Google Analytics and start looking deeper into your data, you'll likely find insights that will challenge your assumptions about how your customers buy.

The important thing to remember is that no one system is right for every merchant. In terms of getting on the path to figuring what attribution model is right for your company, it's critical to look at your data and apply it to different models within the Model Comparison Tool.

Keep in mind that attribution is more of a process than a destination. It's essential to test, refine and implement your process slowly and carefully to ensure that it's the right model for your goals and your different marketing strategies.

For help in applying the Model Comparison Tool and other Google Analytics tools to your marketing channels, reach out to our Paid Marketing team.