Last Updated on 2020-10-12 by ppcguybklyn
When we are checking our Google Ads accounts, we would often see that the number of conversions of the previous month continues to grow after moving on to the next month. Sometimes while doing account audits, we might see the conversion number not being an integer but instead has a decimal number like 0.5. Why would the system add conversions back to the previous month? Isn’t conversion either 0 or 1? Why would there be decimal numbers? All of these questions are related to the conversion window and attribution models of the PPC account. In this article I will be introducing these two concepts and some basic applications.
Conversion window
We could set up our Google Ads accounts to track the conversion actions that have taken place several days after the audience clicks or views the ads to the date when the click or view landed. The time frame that we wish the system would be able to track conversions days after the clicks or views is called conversion window. The default conversion window of Google Ads is 30 days, meaning the system could track conversions actions up to 30 days after clicking or viewing the ads.
Conversion window is closely related to the buying behavior of the target audience. Not everyone would make a purchase right after viewing the ads or landing pages. Some would even decide not to purchase after adding the product to the shopping cart. If the audience makes a purchase several days after seeing our messages through our ads, we could attribute this sale to the ad that the audience first clicked on rather than attribute to the traffic source they use the day they decide to buy. Because the key driver of the purchase is the impression we make when the audience first sees the ad and landing page.
Conversion window’s effect on conversion tracking
The biggest effect on conversion tracking from the conversion window is that Google will continue to add conversion within the conversion window back to the date of the click after concluding a month. Therefore, it is very common to see that even though it is the middle of a month, the conversion number of the previous month is still growing. This is why we don’t necessarily have to run the performance report immediately after the previous month ends. We could wait a few days or even a week to let the conversion window do its job and track more conversions.
Adjust the conversion window based on the behavior of your audience
The conversion window of Google Ads is really flexible. We could adjust the length of the conversion window based on different behaviors from the audience. This can be achieved utilizing the attribution report of Google Ads, combined with basic consumer behavior analysis from Google Analytics to understand the time frame of conversion actions of the customers. Then use this data to set the appropriate conversion window.
Attribution models
Attribution models are different ways our system credit conversions. When our audience click or interact more than once with our ads before they convert, we would need to rely on attribution models to determine which keyword could take the credit for the conversion. What makes attribution models so important is that, in most cases, conversion rate and cost per conversion are our most important metrics. Therefore, how the system attributes conversions to different keywords and ads will determine the priority and approach we adjust our ad campaigns and ad copies. The current attribution models that are available on Google Ads are the following:
Last click
This model gives all the credits of the conversion to the ad and its keyword that the user last clicks on. This is the most straightforward attribution model and also the default model on Google Ads. Using the last click model means we believe the keywords that our customers last interact with are what drive them to make a purchase. Therefore the keywords should take full credit of the conversion. If the last click keywords and ads before our customer convert are the most valuable things to us, then we could use this attribution model.
First click
On the contract to the last click model, this is an attribution model that gives all the credits of the conversion to the first ad and keywords that the customer interacts with. If we believe the final decision of the conversion is determined by the first interaction and impression or the ads, then we could consider using this model.
Last click and first click are the two simplest attribution models. But most of the time the credits of the conversions cannot be solely attributed to the first or the last click. That’s because during the process of a purchase from a customer, every interaction with our ads can play different roles that affect the final decision. Therefore, if we want to further determine the credits of every keyword and ad that are on the conversion path, we could choose to use the following attribution models. The thing these models have in common is that they will all attribute the credit of conversion with their according ratios. This is why when we are using these models we will see non-integer conversion numbers in our system. The higher the attributed conversion number is, the more important the keyword is in the model we are using.
Linear
As the name suggests, this model attributes conversion evenly to all of the interaction on the conversion path. For example, if a customer interacted with the ads from keyword A and keyword B before buying the product, then these two keywords will be credited with 50% of the conversion. The system will show these keywords each has 0.5 conversion. Similarly, if the customer interacted with 4 keywords before making a purchase, then each keyword will be attributed with 0.25 conversion.
Time decay
If we like the concept of the last click model, but we don’t want to deny the credit of the other keywords on the conversion path, then we could use the time decay attribution model. This is a model that is more advanced than linear. The closer the interaction is to the time of the conversion, the more credit would be attributed by this model. The actual ratio of attribution would be based on the number of clicks and interactions on the conversion path.
Position based
If we believe that the first and last clicks are equally important when it comes to determine the customer’s purchase decision, but the rest of the interactions of the conversion path should also be attributed with the conversion credit, then we could use this model. It will attribute 80% of the conversion credit to the first and last clicks and 20% to the rest of the interactions.
Data-driven
The difference between this model and other models is that it will use the account’s data to calculate the actual credit of each interaction on the conversion path, then distributes the credit accordingly. One thing that needs to be aware of is that our account must have sufficient data to use this attribution model. It must have at least 15,000 clicks on Google search and has at least 600 conversions in the past 30 days. Therefore, accounts with low budget or low conversion rate cannot use this attribution model.
Attribution Report on Google Ads
Because it could get complicated when using and analyzing with attribution models, Google Ads offers a specific reporting tool to help us monitor the performance of the account under different attribution models. If we are uncertain of the current attribution model, we could use this tool to compare it with different models. Attribution report also provides data on conversion time so that we could know when the majority of our conversions took place and use this info to determine the best conversion window.
Best practice of attribution models
In most cases I would recommend using the position based attribution model because although it emphasizes on the first and last clicks, it still credits every click on interaction on the conversion path. But it doesn’t mean this is the best attribution model. Our reason behind picking an attribution model should be based on our understanding of our customers. Utilize attribution reports to know the conversion time frame of the customers and use Google Analytics and other tools to understand their conversion patterns, then determine the best attribution model based on this info. Only by doing so can we obtain the most accurate conversion data that can be used to optimize our ad campaigns.
Conclusion
Although conversion window and attribution models can be complicated, it would be extremely helpful for ad optimizations once we further understand the concepts. When operating on Google Ads, a data-driven platform, our best option of making the most profitable decision for our account is to obtain the most useful data by picking the most suitable conversion window and attribution model.
