In this article, I will provide an overview of what Google Ads attributions are and explain data-driven attributions.
Over 50% of all web traffic comes from mobile devices such as smartphones and tablets, which increases year by year. An easy way to find this data on your website is by going to Google Analytics > Mobile > Overview. Users are using more than 1 device to do their product research and purchases.
This new behavior creates a fragmented conversion pattern. Over 90% of users switch between devices to complete a purchase. And over 38% of customers will have interacted with a wide set of platforms. Meaning that they might have started seeing your product on Google Shopping, viewed an ad on display, and clicked on a text ad. So it is now very important that we take each platform into consideration.
Let’s look at an example. When you buy a product, chances are you didn’t use just 1 device or in 1 sitting. Buying a product takes multiple days, research, time, and consideration. Your path before buying a product may be something like below. On Sunday you started your search with a generic search term. On Tuesday your search was more relevant. On Wednesday you found a specific brand. And on Friday you bought your product.
The user journey is a lot more fragmented, different days, times, devices, keywords, advertising areas all leading up to a purchase. Many advertisers are focused on a small subset of consumer paths. They are not considering all the different interactions that lead up to the conversion. Google Ads attributions allow us to see all conversions beyond the last click model.
To make attributions work we need to make sure we track all types of conversion that are important to your business. This can be telephone calls, email link click, contact forms, quote requests, purchases, and so on.
Focus on all conversion types and not just the last click model.
Use the data to improve your advertising goals such as bidding, targeting, and budget decisions.
Make sure you understand what happens after a click
- User completes a search
- User views your ad
- User clicks on your ad
- User arrives on your website
- Do you know what actions your users take on your website?
Make sure you have enabled conversion tracking. You can check this by going to Google Ads > Tools > Conversions
What Is An Attribution
An attribution model is a rule or set of rules, that determines how credits are assigned to the journey along the conversion path. Taking our previous example, we can see that the first interaction started on a Sunday and the last interaction was on Friday where the user bought a product, wherein between we had other interactions.
There are 6 different click interaction models.
- Last click
- Is a static and the last interaction model where 100% of the credit goes before a purchase is made and is the default model within Google Ads.
- First click
- Is a static and first interaction model where 100% of the credit goes to the first click.
- Gives each individual interaction an equal amount of credit.
- Time Decay
- Gives more credits over time in ascending order and provides to the most recent conversion interaction the most credits.
- Assign an equal amount of credit to the first and last click and in-between will have an equal amount of credits.
- It looks to assign credits to individual keywords along the conversion path, depending on the impact it had on the conversion.
Avoid looking at the last click model only and start using all other models. By doing so, we will be able to increase our conversions on average by keeping a similar cost per conversion.
What is Data-Driven attribution? (DDA)
Using all your account data with Google’s machine learning capabilities. Assigns fractional credit, based on the actual contribution of each keyword across the user path. Which learns and adjusts over time. For example: If seasons change and you get a lot more data and conversion cycles change over time, it will adjust its data automatically. For this, to work, you need at least 600 conversions (conversion level) + 15000 clicks in the last 30 days in your Google Shopping campaign (or Search).
Other Conversion Models
What if you do not have enough conversions for the DDA?
Then you can use any of the other conversion models. Not considering the first or last click model as this is not how users are searching anymore, we can look at the other 3 models which are: linear, time decay, or the position based.
Choosing a model that fits your business goals.
Conservative growth strategy
Time Decay: This model is similar to the last click model and is a great option for a conservative growth strategy. This does not give as much value to the early stage clicks.
This model gives 40% equal credit to the first and last click. And the remaining 20% divided into all the clicks that happen in between. Because you are giving equal credit to the first interaction and the last conversion click you are more focused on growth as you are also looking at the first interaction.
No direct goals
Linear based attribution model gives equal credit to each individual click and is a great model if you do not know where you fit if you have a new account and you do not know which growth strategy to use.
Enabling The Right Attribution Model
Log in to Google Ads and go to Tools (top right corner) and choose conversions.
Click on the conversion tracking that you want to update the model for and click on edit settings.
Click on attribution model and choose your model that fits your goal.
Concerns and Misconceptions
I changed my conversion model and sales dropped.
They are not lost or gone, they are simply changed to the model you have chosen. When you compare data before you changed the conversion model and after, the comparison will be based on the last click model before, to the new model after. It is important to remember to compare more accurate fresh data after you changed the model.
The conversions will start to shift between campaigns, ad groups, keywords, search terms.
You will see an initial drop in conversion on the same day you changed the settings and a couple of days after. The reason for this change in data is because the non-last click models are sending conversion credit back in time. Which means the conversion you see dropped today, is sent back to the preceding days where the first click model happened and depending on the conversion window. Make sure you have used a long enough new conversion history to see the conversion data to be the same again.
Important to know:
Other non-last click models can only be seen in Google Ads. The data you see in Google Analytics will only be the last click model.
Source: This article is written using Google Partners resources.