Analytics

Has Google really solved the attribution problem we face as marketers with their latest machine-learning-powered Google Attribution? Find out more in this post.

In May 2017, Google revealed substantial updates to Analytics, AdWords and DoubleClick. These updates are Google’s attempt at solving a few issues that advertisers may face when measuring the success of a campaign. Google publicly announced that this year they will “solve the attribution problem” with the introduction of a new attribution model called “Google Attribution”.

In this blog post, you’ll learn more about the new Google Attribution model, how it works and what this means for us as digital marketers.

First of All, What is Attribution?

Attribution is a set of rules that decides which channel is associated with a particular sale.

Last Click Attribution is divisive in natureYou can understand how vital this could be when working in a digital marketing team. For example, attribution is beneficial if you are in charge of social media, which can be difficult to measure. The value of your individual contribution could be based on the amount of revenue attributed to your specific channel. Consequently, this could lead to ‘discussions’ over who gets the credit for a large order. Currently, there are seven attribution models:

  • Last Interaction (Last-Click)
  • Last Non-Direct Click
  • Last AdWords Click
  • First Interaction
  • Linear
  • Time Decay
  • Position Based

Learn more about the different attribution models.

The default attribution model set on all Google Adwords accounts is Last Interaction.

What is Last Interaction Attribution?

This is the default attribution model. It attributes 100% of the conversion or sales value to the last channel with which the customer interacted before taking action.

So, if a user clicks through an ad then returns the next day directly to the website and converts, the paid ad would not get any credit for the conversion as it was not the last interaction.

Limitations of the Last Interaction Attribution Model

The Last-Click model simple “didn’t cut it”, as emphatically declared by Bill Kee, for the following reasons:

  • It completely ignores how other channels could have assisted in a sale
  • It doesn’t account for multiple devices being used such as mobile, laptop, tablet, etc.
  • You have limited knowledge of what touch points your customer used prior to the sale
  • Lack of integration with ad tools and ad networks does not give a complete picture

Google solving the attribution problem

Take the following below example of someone purchasing a TV.

Let’s say that the user took the following path:

  1. Clicked on a search ad via their mobile
  2. Viewed a display ad on their tablet
  3. Clicked on a Shopping ad on their laptop and then finally made the purchase

In this customer’s journey, which action or touch point should receive the credit for the sale?

Should it be:

  • The initial search ad as this is what brought the user to the website?
  • The display ad as it reminded the user of the TV they looked at?
  • The shopping ad as this was the last interaction before the purchase?

If you were to attribute some credit to all the different touch points, how could this be done fairly?

Now, Let’s Look at a Real-Life Example

In the real world, a typical path is not as simple as Search > Display > Shopping  > Conversion.

Take a supplier of luxury goods –  the typical conversion path is not going to be Search > Display > Shopping  > Conversion. The decision to purchase could span over many weeks or months and will have multiple touch-points along the way that will all attribute to the conversion.

In this example, the conversion paths are far from typical:

Attribution Problem Example Real-life Scenario

Different people tend to take different paths to a conversion and by simply attributing all the credit to the last interaction, you will be missing out on all the valuable touch-points along the way.

This is where Google Attribution comes in – this new model aims to solve these types of questions using machine learning.

So, What is This New “Google Attribution” Model?

Google attribution measures the impact of each marketing touch point across multiple channels, across multiple devices. And it makes it super easy to take action. Taking advantage of data already available in tools like AdWords, Google Analytics and DoubleClick Search – Google Attribution for the first time makes it easy to get a single view of the path to purchase… And best of all, it will be available to all advertisers, at no cost.” – Bill Kee, Google

How Does it Work?

  • Google Attribution will provide a clearer understanding of the customer journey
  • It measures the impact of each marketing touch point across channels and multiple devices
  • It automatically makes use of historical data to drive decision making

All of these enhancements are made possible through advancements in machine learning. This form of artificial intelligence (AI) allows computer programs to learn without additional instructions or code. This is how Google visualises the process in the new model.

Google Attribution model

Source: Google

So essentially, without any additional work, Google will pull all the data from Analytics and Adwords together, analyse the performance of multiple channels and devices and take action without you having to do anything.

How Will Google Attribution Help Me?

This advancement in technology means that any credit to sales activity can be given fairly.

This approach looks at the whole picture and brings data from often overlooked channels such as:

  • Organic search
  • Email
  • Affiliates
  • Social
  • Direct visits

In addition, there will also be an enterprise version called Google Attribution 360 for larger advertisers using DoubleClick Campaign Manager.

Currently, the data-driven attribution model available in Adwords requires an account to have “15,000 clicks and a conversion action must have at least 600 conversions within 30 days”. So this is only really applicable to big spending accounts, but the new Google Attribution model doesn’t have any barriers and is available to everybody.

Google Attribution will be free for all

As a Google Premier Partner, we have access to the beta and our Head of Insights, Jonathan Ellins is currently running tests across a number of accounts. So far, the accuracy seems to have improved, however it’s still very early days. What is promising is that Google claims the new data driven model will be “5% to 15% more accurate”.

Conclusion

In conclusion, Google recognises that there has been a problem and looks to be addressing it. Since we live with multiple devices in a more connected world, it makes sense to have a smarter attribution model. The entire success of a marketing campaign cannot be viewed by the last interaction prior to a sale or enquiry.

Google Attribution is data-driven attribution and allows you to value all conversion data to calculate how much credit each channel deserves. The great thing about it is that it will continually adapt according to changes in your data through its algorithm.

Furthermore, this could provide a more accurate view of your marketing. You would be able to allocate budgets according to which channel has the most impact in the customer’s journey.

Google claims that advertisers using this new attribution model will be able to make better marketing decisions. As a result, you could gain a better return on your ad campaigns without additional spend.

If you need help understanding this new update, you can speak to one of the experts.

Leave a Reply