Guide to Attribution Modelling in Google Ads (Adwords)

In the context of Google Ads (formerly Google Adwords), attribution modelling is the process of understanding the role of various keywords, ad groups and campaigns in:

  • Initiating conversion actions
  • Assisting conversion actions
  • Completing conversion actions

Attribution modelling is a very broad topic.

This article talks only about understanding and using attribution modelling in the context of Google Ads.

To understand the very basics of attribution modelling and how it works in Google Analytics, check out this article: Beginners Guide to Google Analytics Attribution Modelling

If you have not already read this article then I would highly recommend that you read it.

It is a good starter guide to understand the various aspects of attribution modelling and can help you greatly in understanding the present article.

Why You Should Use Attribution Modelling in Google Ads?

You should use attribution modelling in Google ads to understand the role of various keywords, ad groups and campaigns in initiating, assisting and completing conversions.

Through attribution modelling you can understand how users search for your products and what they do before they make a purchase.

The biggest insight that you can get from attribution modelling is that you can determine the most effective keywords, ad groups and campaigns for investment.

Note: Attribution modelling in Google Ads is available only for clicks on Search Network ads and Shopping ads on google.com. It is not available for interactions with Display Network ads.

Say Goodbye to ‘Last Ad Click Keywords’ Optimization

The advent of attribution modelling has fundamentally changed the way we look at the data for good.

Up to now we have been managing PPC campaigns by bidding on ‘last ad click keywords’ (keywords which completed the sales).

This has to change because it is sub-optimal way of optimizing a campaign.

To get optimum results, you also need to invest in keywords which initiated or assisted in sales.

Different keywords (first ad click keywords, middle ad click keywords and last ad click keywords) work together to create a sale.

Yet we remain focus only on the last click keywords.

Not good.

It is quite common for users to refine their search queries (on search engines like google.com) to get the product they are looking for.

And it is also quite common for users to return to a website via a branded keyword (as it is easy to remember) before making a purchase.

Since by default, Google Analytics attributes conversions to the last non-direct clicks, branded keywords end up getting tons of undue credit for sales.

Consequently branded keywords campaigns seem to perform so well in PPC.

But these campaigns are still not producing optimum results because our whole optimization efforts are directed towards last click keywords and we now know, that different keywords work together to create a sale.

If you keep adding more of one unit of production to a productive process while keeping all others units constant, you will at some point produce lower per unit returns.

This is known as the law of diminishing returns.

So in the case of PPC, if you keep optimizing for last click keywords (in our case branded keywords) while ignoring first and middle clicks keywords (collectively known as assist keywords) you will at some point produce lower per unit returns.

What that means, your cost per acquisition at some point will start rising and your profit on sales will start declining.

Then the only way, to remain within your CPA targets is by tweaking (add, pause, delete, change bids) last click keywords.

But this is sub-optimal way of optimizing a PPC campaign, as you are optimizing only the bottom part of the conversion funnel.

So what is the solution?

In order to strengthen your PPC campaigns, you also need to bid on keywords and search queries that initiated and assisted conversions (conversion actions).

In this way you can stay away from the point of diminishing returns and remain within your CPA targets much longer.

How to Find Keywords and Search Queries that Initiated and Assisted Conversions via Google Analytics

Identify keywords and search queries in your Google Analytics multi-channel funnel reports that are:

  1. High performing conversion initiators (i.e. the top keywords/search queries which initiated conversions)
  2. High performing conversion assisters  (i.e. the top keywords/search queries which assisted conversions).

Follow the steps below:

Step-1: Navigate to ‘Assisted Conversions Report’ (‘Conversions’ > ‘Multi Channel Funnels’) in your main Google Analytics view.

Step-2: From the ‘Conversions’ drop down menu, deselect all conversion types except ‘ecommerce transaction’:

select-conversion-type

Step-3: Click on the ‘Google Ads’ button:

Step-4: Click on ‘Assisting Interactions Analysis’ tab in the ‘Explorer’ window:

Step-5: Click on ‘Google Ads Keyword’ as primary dimension:

Step-6: Set the date range to the last 3 months:

Step-7: Sort the report by ‘Assisted Conversion Value’ column to find high performing conversions assisting keywords:

Step-8: Compare all the high performing assisting keywords against the keywords you are currently bidding on in your PPC campaigns. Add the keywords which are missing, to your PPC campaigns and then bid on them.

Step-9: Select ‘Search Query’ as primary dimension and then sort the report by ‘Assisted Conversion Value’ column to find high performing conversions assisting search queries:

Step-10: Compare all the high performing assisting search queries against the keywords you are currently bidding on in your PPC campaigns. Add the search queries which are missing (as keywords), to your PPC campaigns and then bid on them.

Step-11: Click on ‘First Interaction Analysis’ tab in the ‘Explorer’ window:

first-interaction-analysis

Step-12: Sort the data by ‘First click Conversion Value’ column to find high performing conversions initiator search queries:

Step-13: Compare all the high performing conversion initiator search queries against the keywords you are currently bidding on in your PPC campaigns. Add the search queries which are missing (as keywords), to your PPC campaigns and then bid on them.

Step-14: Click on ‘Google Ads Keyword’ as primary dimension and then sort the data by ‘First click Conversion Value’ column to find high performing conversions initiator keywords:

Step-15: Compare all the high performing conversion initiator keywords against the keywords you are currently bidding on in your PPC campaigns. Add the keywords which are missing, to your PPC campaigns and then bid on them.

Say Goodbye to ‘Last Ad Click CPA Optimization’

The ‘Cost Per Acquisition‘ (also known as ‘CPA’ or ‘Cost per conversion’) that you see in your Google Ads report is not your actual cost per acquisition:

It is the ‘cost per last ad click conversion’.

So if you ignore first and middle click keywords and optimize PPC campaigns on the basis of ‘cost per last ad click’ conversions then you won’t get optimal results and may even lose money.

That is because if a keyword is not completing a sale, it may be initiating a sale or assisting a sale (always remember that).

And if you stop bidding on the keyword because its ‘cost per last click ad conversion’ (the so called CPA reported by Google Ads) is too high or the keyword is not directly completing any conversion then you could lose money.

Similarly the conversions reported by Google Ads are last ad click conversions.

So if you overlook the role of prior keywords in the conversion process then you are missing the big picture.

Introduction to Ad Impressions

Every time your ad is served on Google Ads Networks (Google.com, YouTube, Partner websites and apps), it is counted as ‘Ad Impression’ by Google ads.

It does not matter whether or not your ad was actually viewed by a user.

As long as your ad is served on Google Ads Networks, each serving is counted as one ad impression.

For example, if you ad was served 10 times on Google Ads Networks then the number of impressions counted for the ad would be 10.

Introduction to Measurable and Non-Measurable Ad Impressions

Measurable ad impressions is the number of times your ad could be measured for viewability.

Non-Measurable ad impressions is the number of times your ad could not be measured for viewability.

Ad Impressions = Measurable Ad Impressions + Non-Measurable Ad Impressions

For example, if your ad got 6 measurable impressions and 4 non-measurable impressions then total ad impressions counted for your ad would be: 6 + 4 = 10

Introduction to Viewable and Non-Viewable Ad Impressions

Viewable ad impressions is the number of times your ad was viewable. An ad is viewable when at least 50% of its area is visible for 1 second for Display Network ads or 2 seconds for video ads.

Non-Viewable ad impressions is the number of times your ad was considered non-viewable. An ad is non-viewable when at least 50% of its area is not visible for 1 second for Display Network ads or 2 seconds for video ads.

Measurable Ad Impressions = Vieweable Ad Impressions + Non-Vieweable Ad Impressions

For example, if your ad got 4 vieweable impressions and 2 non-vieweable impressions then total measurable ad impressions counted for your ad would be: 4 + 2 = 6

Summary:

Ad Impressions = Measurable Ad Impressions + Non-Measurable Ad Impressions

Ad Impressions= (Vieweable Ad Impressions + Non-Vieweable Ad Impressions) + Non-Measurable Ad Impressions

Introduction to Ad Click

The event of clicking on an ad is called ad click.

A user can not click on an ad without viewing it.

So an ad click is always followed by one or more ad impressions.

Introduction to Views (or Video Views)

A View (or Video View) is counted everytime a user:

  • watched 30 seconds or less of your video ad.
  • clicked on call to action overlay, cards or companion banner.
  • clicked on the thumbnail for video discovery ads.

Note: Video views do not include video ad impressions or thumbnail impressions for video discovery ads.

Introduction to Ad Engagement

An engagement is counted every time a user:

  • expand a Lightbox ad or a card on a video ad
  • view or click on a Showcase Shopping ad.

Note: Ad engagement generally does not include: ad impressions, ad clicks and video views.

Introduction to Ad Interactions

An interaction is the main action associated with an ad format.

Interactions can include: ad clicks, ad engagement or video views.

Example-1:

For ‘Search Campaigns‘, ad interactions include ad clicks.

So if a search campaign has got 100 ad clicks in the last 30 days then Google will report 100 ‘ad interactions’.

Example-2:

For ‘Display Campaigns‘, ad interactions can include both ad clicks and ad engagement.

So if a display campaign has got 53 ad clicks and 27 ad engagements in the last 30 days then Google will report 80 ‘ad interactions’.

These 80 interactions is the sum of 53 ad clicks and 27 ad engagements

Example-3:

For ‘Shopping Campaign‘, ad interactions can include both ad clicks and ad engagement.

Example-4:

For ‘Video Campaigns‘, ad interactions include ‘video views’.

So if a video campaign has got 100 views in the last 30 days then Google will report 100 ‘ad interactions’.

 

Note(1): Ad interactions do not include ad impressions.

Note(2): A video ad impression is different than a video view.

Introduction to Conversion Action

In the context of Google Ads, conversions are called ‘conversion actions’.

A conversion is a website goal or objective.

Following are examples of conversion actions:

  • purchase
  • lead
  • signup
  • viewing of a key page
  • phone call
  • app install
  • store visit etc.

In order to view or create conversion actions in your Google Ads account, click on the ‘Tools’ tab and then click on the ‘Conversions’ link under the ‘Measurement’ section:

 

Types of Conversion Actions

In the context of Google Ads, there are four types of conversion actions.

You can see these different types of conversion actions by clicking on the + button:

 

#1 Website Conversion Actions – the conversion actions recorded on your website (like purchase, leads, sign up, viewing of a key page etc).

#2 App Conversion Actions – the conversion actions recorded on your mobile app (like mobile app installs, in-app actions like purchase).

#3 Phone Calls Conversion Actions – Following conversion actions are recorded and reported as phone calls conversion actions by Google Ads:

  1. Calls to a phone number in your ads
  2. Calls to a phone number on your website.
  3. Clicks on a phone number on your mobile website.

#4 Imported Conversion Actions – As the name suggest, such type of conversion actions are imported into your Google Ads account from other accounts like:

  • Google Analytics
  • Firebase
  • Third Party App Analytics
  • Salesforce
  • Other data sources or CRMs

Note: Attribution modelling in Google Ads does not take app conversions and in-store conversions into account.

Categories of Conversion Actions

There are three broad categories of conversions actions (or conversions):

#1 Click through Conversion Action (also known as click through conversion)

#2 View through Conversion Action (also known as view through conversion)

#3 Cross device Conversion Action (also known as cross device conversion)

A ‘click through conversion’ is counted when a user clicks on a Google ad and then complete a conversion on your website / mobile app.

A ‘view through conversion’ is counted when a user completes a conversion on your website / mobile app after viewing (but not interacting) a Google display network ad or Google Video ad.

A ‘Cross device conversion’ is counted when a user interacts with your ad on one device/browser but completes a conversion on another device/browser.

Introduction to Conversion Paths in Google Ads

In the context of Google Ads Attribution Modelling, a conversion path is the sequence of ad interactions with your search network ads and/or shopping ads (on google.com) during the 30 to 90 days period that lead to a conversion action.

It is important to note that, in the context of Google Ads attribution modelling, a conversion path is just made of ad clicks and/or ad impressions.

The conversion path does not take the role of other marketing channels (like ‘organic search’, ‘social media’, ‘email’ etc) into account.

Introduction to ‘History Window’ in Google Ads Attribution Modelling

The period of 30 to 90 days prior to conversion action is called the ‘History Window’.

In Google Ads, the history window can have four possible values:

  1. Default (the conversion window you defined while setting up a conversion action)
  2. 30 days
  3. 60 days
  4. 90 days.

The ad clicks and ad impressions which were recorded within the history window are considered for getting credit for a conversion action by Google ads.

The History Window is available in almost all attribution modelling reports in Google Ads.

To see the history window, follow the steps below:

Step-1: Click on the ‘Tools’ tab and then click on the ‘Search Attribution‘ link under the ‘Measurement’ section:

Step-2: Click on the ‘Attribution Modelling’ link in the left hand side navigation. You should then be able to see the ‘History Window’ drop down menu on the right:

Introduction to Conversion Window

Google ads refer to ‘History Window’ as ‘Conversion Window’ while setting up conversion actions.

When you are creating or editing conversion actions in Google ads, you get the option to set the conversion window.

Types of Conversion Windows

There are two types of Conversion Window:

  1. #1 Click through Conversion Window (usually refer to as just ‘conversion window’).
  2. #2 View through Conversion Window

Click through conversion window refer to the time period within which, an ad click is considered for getting credit for a conversion action by Google Ads.

View through conversion window refer to the time period within which, an ad impression is considered for getting credit for a conversion action by Google Ads.

How to View and Edit Click Through/View Through Conversion Window in Google Ads

To view and/or edit the click through/view through conversion window, follow the steps below:

Step-1: Click on the ‘Tools’ tab and then click on the ‘Conversions’ link under the ‘Measurement’ section:

Step-2: Click on the name of an existing conversion action:

You should now be able to see conversion window and/or view through conversion conversion window for a particular conversion action:

Step-3: Click on the ‘Edit Settings’ link and then click on the ‘Conversion Window’ row in order to edit the click through conversion window:

You can set the conversion window to any time period that best match with your sales cycle:

In order to determine, how long it takes your customers (who see and clicks your ads) to complete a conversion on your website, segment your reports by “Days to conversion.”

Likewise, you can edit the view through conversion window while creating or editing a conversion action in Google Ads:

Introduction to Attribution Models in Google Ads

An attribution model is a set of rules that determine how the credit for conversions should be attributed to various keywords, ad groups and campaigns in Google Ads.

Attribution models in Google Ads can be broadly classified into two categories:

#1 Rule Based Attribution models

#2 Algorithmic attribution models

Rule Based Attribution Models

Rule Based Attribution models assign conversion credits to ad clicks based on some predefined rules.

Following are the examples of rule based attribution models available in Google Ads:

  1. Last Click
  2. First Click
  3. Linear
  4. Time Decay
  5. Position-based

Algorithmic Attribution Models

Algorithmic attribution models assign conversion credits to ad clicks based on an algorithm.

The ‘Data Driven Attribution Model’ is an algorithmic attribution model offered by Google Ads.

Last Click Attribution Model

This attribution model gives all the credit for conversion to the last clicked ad and corresponding keyword.

The default attribution model used by Google Ads is the ‘Last clickmodel.

So by default the last clicked ad (and corresponding keyword) in a conversion path gets all the credit for conversion in Google Ads.

Use this model, if least amount of buying consideration is involved.

For example, if you are a FMCG (fast moving consumer goods) company like ‘Tesco’ / walmart and you sell products which involves the least amount of consideration by a buyer (like purchasing toothpaste), then you can use the last click attribution model.

This is because you do not need to assign more conversion credit to the first and middle interactions in your conversion path.

Get the FREE E-Book (52 Pages)

First Click Attribution Model

This attribution model gives all the credit for conversion to the first clicked ad and corresponding keyword.

In other words, the first clicked ad (and corresponding keyword) in a conversion path gets all the credit for conversion.

Use the first click model, if brand building/brand awareness is very important for you.

If you are new player in your niche then you may need more brand awareness than your established competitors.

Consequently your advertising goals should be more ‘brand building’ centric. 

So you may need to assign more conversion credit to ad clicks, which initiated the conversion process.

In that case you use the First click attribution model.  

Linear Attribution Model

This attribution model gives equal credit for conversion to all the ad clicks in a conversion path.

Use this attribution model, if you have a business model where each interaction with your customers is equally important for your conversions.

For example, if you provide customers support service then each interaction with your customers is equally important for you.

In that case, use the linear model.

Time Decay Attribution Model

This attribution model gives more credit to the ad clicks that happened closer in time to the conversion.

The conversion credit is distributed using a 7 day half life.

What that means, the ad click which is recorded 8 days before a conversion, gets half as much credit as the ad click recorded 1 day before the conversion.

Use the time decay model, if you are running time sensitive promotional campaigns.

If you want to understand the buying behavior of your customers during a promotional campaign then you would want to assign more credits to the ad clicks which occurred closest in time to conversions ,as they are more relevant than the ad clicks which occurred further in the past.

In that case, use the Time Decay Model.

In the absence of data driven attribution model, I prefer to use the time decay model.

Position Based Attribution Model

This attribution model gives 40% conversion credit to first clicked ad (and corresponding keyword) and 40% conversion credit to last clicked ad (and corresponding keyword).

The remaining 20% conversion credit is equally distributed among all other ad clicks in a conversion path.

If you have a business model or advertising objectives, which value first and last customer interactions more than the middle interactions, then use the Position based attribution model.

Data Driven Attribution (DDA) Model

The ‘Data Driven Attribution Model’ is an algorithmic attribution model offered by Google Ads.

The data driven attribution model uses a predictive algorithm (algorithm used to predict the users’ probability of making a purchase or completing some other conversion) to assign or reassign conversion credit to various ‘ad interactions’ in a conversion path according to the most recent conversion data.

It is a conversion probability model.

It does not distribute conversion credit just on the basis of ad interactions’ position.

In other words ‘data driven attribution model’ is not a rule/position based model.

Use the data driven attribution model, if your Google ads account is consistently getting:

  • 15000 or more ad clicks each month on Google Search.
  • 600 or more conversion actions each month.

Attribution Models example

Let us suppose you own a business based which provides ‘car insurance’.

Let us also suppose that you run Google Ads to advertise your business.

Now consider the following conversion path:

A prospect find your website for the first time by clicking on your ad which was triggered for the keyword ‘car insurance UK’. He did not make any purchase for some reason.

The prospect later returned to your website by clicking on your ad which was triggered for the keyword ‘cheap car insurance’. He again did not make any purchase.

The prospect later returned to your website for the third time by clicking on your ad which was triggered for the keyword ‘buy car insurance UK’. He again did not make any purchase.

The prospect returned to your website once again by clicking on your ad which was triggered for the keyword ‘car insurance Southampton’. This time the prospect purchased car insurance on your website.

So the keyword conversion path would look like the one below:

car insurance UK > cheap car insurance > buy car insurance UK > car insurance Southampton

Now under the “Last click attribution model”, the last keyword, “car insurance Southampton” would receive 100% of the credit for the conversion.
Under the “First click attribution model”, the first keyword, “car insurance UK” would receive 100% of the credit for the conversion.

Under the “Linear attribution model”, each keyword would receive equal credit (25% each) for the conversion.

Under the ‘Time decay model’, the keyword “car insurance Southampton” would receive maximum credit for the conversion (as it was searched closest to the conversion) followed by the keyword  ‘buy car insurance UK’, ‘cheap car insurance’ and ‘car insurance UK’. The keyword ‘car insurance UK’ would receive least credit for the conversion as it was searched farthest from the conversion.

Under the “Position-based attribution model”, the keywords “car insurance UK” and “car insurance Southampton” would each receive 40% conversion credit, while “cheap car insuranceand “buy car insurance UK” keywords would each receive 10% conversion credit.

Under the “Data driven attribution model”, each keyword would receive conversion credit according to their contribution in a conversion path.

The keyword which assisted / influenced the conversion the most, gets the maximum credit for conversion, regardless of it being a first click , last click or middle click. All other keywords would get credit in proportion to their contribution in the conversion path. The conversion credits are assigned using a conversion probability model which uses a predictive algorithm.

How to View/Change/Set the Attribution Model Currently Being Used in Google Ads

While creating / editing a conversion action, you can get the option to see or edit the attribution model currently being used:

The attribution model that you select while editing/creating a conversion action tells Google ads, how to attribute conversions for the conversion action.

Note: The attribution model that you select only affect the conversion action to which it is applied and it only affect your search network and shopping conversion data. The selected attribution model does not affect your Display network, Mobile App and phone call conversion data.

For step by step instructions, read this article: How to change Attribution Model in Google Ads (Google Adwords)

Google Ads Search Funnels

From the chart above we can conclude that Google Ads search funnel is a type of single channel paid search funnel.

There can be two types of Google Ads search funnels:

#1 Google Ads goal search funnel – it is a sequence of interactions ad clicks and ad impressions that lead up to a non-transactional goal.

#2 Google Ads sales search funnel – it is a sequence of ad clicks and ad impressions that lead up to a transaction goal.

As such, a traditional MMM model is not suitable for carrying out digital marketing mix modelling aka attribution modelling.

Other Articles Related to Attribution Modelling in Google Ads

  1. Which Attribution Model to use in Google Ads (Adwords)?
  2. Cross device attribution reports in Google Ads (AdWords)
  3. How to change Attribution Model in Google Ads (Adwords)
  4. Data Driven Attribution Model in Google Ads (Adwords)
  5. Setting up Data Driven Attribution Model in Google Ads (Adwords)
  6. Store visit conversion tracking in Google Ads (AdWords)
  7. Understanding the Analytics behind Google Ads (Adwords)
  8. Cross Account Conversion Tracking in Google Ads (Adwords)
  9. Understanding View-Through conversions In Google Ads (Adwords)
  10. Assisted Conversions Report in Google Ads (Adwords)
  11. Understanding Conversion Paths in Google Ads (Adwords)
  12. Attribution Modelling Reports in Google Ads (Adwords)
  13. How to test Attribution Model in Google Ads (Adwords)

Learn about the Google Analytics Usage Trends Tool

The Google Analytics usage trend is a new tool which is used to visualise trends in your Google Analytics data and to perform trend analysis.


Do you want to Learn Web Analytics in 4 Weeks?

  • Learn and Master Web Analytics, Conversion Optimization & Google Analytics from Industry Expert in 4 weeks.
  • Lifetime access to the course + Lifelong FREE course updates.
  • New study material added every few months (lifelong learning).
  • Up to date training material.
  • Most exhaustive course on Google Analytics on the internet.
  • Hundreds of Assessments to test your learning.
  • Your 24/7, 365 days a year reference source.
  • Learn at your own pace and from any place.
  • Risk Free with 30 days 100% Money Back Guarantee.

Take your Analytics knowledge to the next level. Checkout my Best Selling Books on Amazon

Maths and Stats for Web Analytics and Conversion Optimization
This expert guide will teach you how to leverage the knowledge of maths and statistics in order to accurately interpret data and take actions, which can quickly improve the bottom-line of your online business.

Master the Essentials of Email Marketing Analytics
This book focuses solely on the ‘analytics’ that power your email marketing optimization program and will help you dramatically reduce your cost per acquisition and increase marketing ROI by tracking the performance of the various KPIs and metrics used for email marketing.

Attribution Modelling in Google Analytics and Beyond
Attribution modelling is the process of determining the most effective marketing channels for investment. This book has been written to help you implement attribution modelling. It will teach you how to leverage the knowledge of attribution modelling in order to allocate marketing budget and understand buying behaviour.

Himanshu Sharma

Certified web analyst and founder of OptimizeSmart.com

My name is Himanshu Sharma and I help businesses find and fix their Google Analytics and conversion issues. If you have any questions or comments please contact me.

  • Over eleven years' experience in SEO, PPC and web analytics
  • Google Analytics certified
  • Google AdWords certified
  • Nominated for Digital Analytics Association Award for Excellence
  • Bachelors degree in Internet Science
  • Founder of OptimizeSmart.com and EventEducation.com

I am also the author of three books:

error: Alert: Content is protected !!