Are You Using the Right Attribution Model in Google Ads?

No comment

Do you know the value of your Google Ads investment?

As marketers, we have what can feel like endless options to utilize our often limited marketing budgets.

How do you know where to invest? PPC is often a low-funnel highly-targeted place to start, but how do you justify the results you are driving? Should you be spending more? Less?

All these questions are crucial to getting the most out of your broader marketing mix. But the answer isn’t always as straightforward as it seems, even in the world of PPC where calculating your ROI is often much easier than most other digital channels (and certainly easier than more traditional marketing channels).

It sounds simple. You have Google Analytics and conversion tracking set up. So you ask yourself, “did my click result in a conversion?”

But many other questions come from this. How much is a conversion worth? How often do your conversions result in revenue? What are your margins?

These are much larger questions that we’ll save for another blog post. The one we’ll focus on here is simply, “who deserves credit for a conversion?” And that brings us to attribution.

What is Attribution?

Attribution is considered by many to be the most important marketing argument. In marketing, attribution is the action of deducing what deserves credit for a conversion. Sounds relatively straightforward, right?

However, things can get messy quickly as sales funnels get longer. For instance, what deserves credit for the conversion in this user path scenario?

Example of a user path with five touchpoints before conversion.Example of a user path with five touchpoints before conversion.

 

Google Ads and Google Analytics both default to an attribution model called last click. So with those defaults in place, as many businesses use, 100% of the credit for the conversion above would go to a direct site visit, despite the heavy influence paid search, paid social, and display played in the nine-step path this user traveled down on their way to converting.

When you lay it out like that, it’s easy to see there are unlimited different combinations of touchpoints that could result in a conversion…so maybe last click doesn’t always fit.

Let’s take a look at the other attribution models available within Google Ads.

Attribution Models

In Google Ads, there are six different attribution models to choose from that distribute credit in different ways. I’ll go through them and discuss when you might want to use each.

Last Click

Last click attribution gives 100% of the credit for a conversion to an ad when it was the last click that led to the conversion (again, this is the default setting).

Visual representation of last click attributionVisual representation of last click attribution

Example: This model is best used when trying to identify which channels are most responsible for bottom of the funnel deal closing.

 

First Click

First click gives 100% of the credit for a conversion to an ad that was the first click in a path that led to a conversion.

Visual representation of first click attributionVisual representation of first click attribution

Example: This model is best used when trying to identify how users are introduced to your brand (new users).

 

Linear

Credit is distributed evenly across all clicks that the user had on their way to a conversion.

Visual representation of linear attributionVisual representation of linear attribution

Example: This model is more useful when you are trying to understand how all your channels are playing a role in a conversion.

 

Time Decay

Credit is weighed by the click’s proximity to the conversion using a seven-day half-life. A click eight days before a conversion gets half the credit of a click one day before the conversion.

Visual representation of time decay attributionVisual representation of time decay attribution

Example: This model is best used when trying to learn what strategies are better brand introducers and what strategies drive the most conversions.

 

Position-Based

Credit is distributed with 40% credit towards the first click and 40% credit towards the last clicks. The remaining 20% is spread evenly between the clicks in the middle.

Visual representation of position based attributionVisual representation of position based attribution

Example: This model is most useful when you value brand introduction and conversion driving the most.

 

 

Data Driven

Credit is shared based on the weight of each click. Clicks that play a more significant role historically in driving conversion earn more weight. This model is specific to each advertiser’s individual conversion data.

Visual representation of data driven attributionVisual representation of data driven attribution

Example: This model is the most scientific. It uses algorithms to weigh conversions, and with enough data, is most useful when trying to understand which channels are playing the largest role in driving conversions.

 

Each model has pros and cons inherently. So how do you pick one that is going to work for you?

Choosing an Attribution Model

In many cases, we are all so used to last click attribution that it’s hard to buck the trend and start fresh with something new, even if we know it is misrepresenting the data. We all have reports to create and people to report to.

With that in mind, it’s important to decide on a model using facts, data, and a clear understanding of your goals and expectations.

Google Analytics has a nifty tool to help.

Model Comparison Tool

In Google Analytics under Conversions > Attribution you will find the Model Comparison Tool.

Screenshot showing where to locate the model comparison tool in Google AnalyticsScreenshot showing where to locate the model comparison tool in Google Analytics

 

This tool allows you to pit up to three different attribution models against one another for the same data set to see how your conversion data changes as you begin to consider different models.

Specifically, you can toggle the tool to just view Google Ads data.

 

Screenshot of the model comparison toolScreenshot of the model comparison tool

 

Start with Last Interaction (i.e. last click) and start comparing different models. Check the “% change in conversion” column to begin to get a picture of how Google Ads performs differently with a different model in place. You’ll start to notice that some campaigns and strategies are affected differently by different models.

Before Making a Switch

There is a lot to consider before making the jump to a new model in Google Ads.

Consider your business goals. What are you trying to get out of Google Ads? If you are investing significant budget into Google Ads, it’s likely you are expecting conversions. But are you limiting yourself in scenarios like the one we outlined at the start?

Review your Top Conversion Paths in Google Analytics to learn what paths users are most often taking.

Screenshot showing where to locate the top conversion paths in Google AnalyticsScreenshot showing where to locate the top conversion paths in Google Analytics

 

When you do change your attribution model in the Conversions section of Google Ads, your data will begin to shift and it will not retroactively update historical data. You’ve most likely been optimizing towards one model for some time, so make sure everyone is on the same page before switching.

Final Thoughts

Attribution is a huge question, and most of the biggest brands and highest spending advertisers struggle with it. There is very likely not a one size fits all answer to “the right” attribution model for you.

But just because the industry default has historically been last click doesn’t mean you have to continue the trend. Weigh your options, look at the data, consider your goals, and pick a model that reflects your strategies and business.

Once you settle on a new, more accurate model, you can begin collecting new data. In theory, the new data will be more representative of your goals, so you should be able to start making optimizations off it that will help improve your overall business KPIs.

The post Are You Using the Right Attribution Model in Google Ads? appeared first on Portent.