Google Ads Conversion Attribution Models
Have you ever noticed that you can have less than 1 conversion within your Google Ads account? Do you see decimal places like .6 conversions appearing in your reports? This is because certain attribution models in Google give partial credit of conversions to campaigns, ad groups, ads, and even keywords.
How Google distributes the conversion value is all based off which attribution model you have your conversion actions set to.
What are attribution models?
Attribution models are the rules or set of rules in which that determines how credit is dispersed based off of various touch points in the sales cycle. I’ll be going over the 6 different attribution models used in Google Ads in this blog: Data-driven, Last click, First click, Linear. Time decay, and Position-based.
Much like the name states the last click attribution model simply gives 100% of the credit to the last click. In Google Ads, that means the credit goes to the last ad and corresponding keyword clicked before the conversion action took place.
This model is helpful when you want to identifying your most bottom of funnel keywords or ads. This is also the default setting for Google Analytics and cannot be changed (unless you invest in Google Analytics 360). If you do chose to use the last click attribution model you will be missing other touch points that helped move a user to convert which could result in pausing or deleting what looks like under performing campaigns, ad groups, ads, or keywords when they really are playing an important role in the sales cycle.
Another simple attribution model, first click gives all of the credit to the first clicked ad and corresponding keyword (if search), ad group, and campaign.
This model is helpful in identifying where users might start their journey in your sales cycle. That said, if you use this model you would lose insights into the true conversion path as all credit would be given to the first click and not necessarily the final conversion driving click.
The linear attribution model identifies all of the touch points that happen throughout a conversion. With all of the touch points identified Google is able to evenly disperse the conversion across each touch point. So, for example, if a user clicks on three ads before finally converting, each would be given .33 of a conversion.
This model is helpful as it gives more insights into the conversion path of a user.
The time decay attribution model gives mores credit to the clicks that happened closer to the conversion. Credit is dispersed based off of a 7-day half life model, credit for a click will be halved every 7 days after the last click to conversion happened. For example, a click that happened 8 days before the conversion occurred will have received half as much credit as a click that happened 1 day before the conversion because 7 days had gone by prior to conversion.
This attribution model gives more credit to the first and last click sources. Google gives 40% to each the first and last clicked ads and distributes the remaining 20% across the other clicks in the path. This method too will lead to partial conversions of less than 1.
In order to enable the data driven attribution model you must have had enough clicks from Google Search (15,000) and conversion events (600) over the last 30 days. Google sets these requirements for the data driven attribution model because it uses your conversion history in order to assign credit to each campaign, ad and keyword in the conversion path. Google compares the click path of users who converted versus users who don’t convert and looks for unique patterns. Through these unique patterns Google is able to identify certain steps in the sales cycle that have a higher impact on whether or not the user will convert. By identifying patterns like these Google is able to give the higher valued events in your sales cycle more credit when the conversion event takes place.
This attribution model is highly recommended if your account meets Google’s requirements. If you drop below 10,000 clicks on Google search or 400 conversions within 30 days you’ll receive an alert letting you know. If your data continues to remain at the low volume over the next 30 days you’ll attribution model will revert back to the Linear model.
Scenario:
You own a flower shop in San Francisco. A customers finds your website after searching the following searches: ‘best flower store in san francisco’, ‘red roses san francisco,’ and finally ‘order red roses online san francisco’. The user then buys a bouquet of red roses online, but which query and corresponding ads and keywords get credit?
- Last click: The query ‘order red roses online san francisco’ and the corresponding ad and keyword would receive all of the credit.
- First click: The query ‘best flower store san francisco’ and the corresponding ad and keyword would receive all of the credit.
- Linear: Each corresponding keyword and ad would receive of the three queries would receive an equal amount of credit (33% or .33 of a conversion)
- Time decay: ‘Order red roses online san francisco’ would receive the most credit of the three because it is closest to the actual conversion event.
- Position based: The corresponding keywords and ads for ‘best flower store san francisco’ and ‘order red roses online san francisco’ would each receive 40% of the credit. The remainder (20%) would be attributed to ‘red roses san francisco.’.
- Data driven: Each corresponding keyword and ad would receive partial credit. The amount of credit depending on which Google considers to be of more or lesser value in the sales cycle.
As I mentioned above Google Analytics uses the Last Click model and cannot be changed unless you invest in Google Analytics 360, which will cost your $150,000/year.
If you do create a goal in Google Analytics and import that into Google Ads as a conversion event you’ll be able to change the attribution model within Google Ads, but will it work if the source of the conversion (Google Analytics) is set to Last Click? The answer is simple – Yes!
Since Google Ads and Analytics both utilize GCLID’s, (Link to a blog if we have one or I’ll write about this) Google Ads will be able to report based off of a different attribution model. The only drawback is that Google Analytics will only report based off of the Last Click model therefore your data will not line up properly if you are pulling from both Google Analytics and Google Ads.
We at Four15 Digital strongly believe in and encourage the use of machine learning in every aspect of your Google Ads account. That is way, if you’re eligible, we highly recommend to use the Data Driven attribution model. This will give you the best conversion insights and allow you (or Four15 Digital) to optimize your account as best as possible. If the Data Driven model is unavailable to you, the Time Decay attribution model is recommended. Time Decay allows you to still track clicks that happened before the final click while dispersing the value in a more strategic way than the Linear model does.
