Whether your customers are shopping online, in-store, or both, this year, it's important to ensure all measurement tools are in place to battle blind spots from cross-device shopping.
Think back to the last time you clicked on an ad. Was it your first time seeing the ad or was it your fifth? Did you make a purchase right then and there, or did you turn to Google for further information? Regardless of how long it took between initial impression to final conversion, credit was applied to each channel that influenced your decision to convert.
Digital advertising relies on accurate data measurement to deliver timely and relevant messages, influence strategic campaign-wide decisions, and guide optimization efforts. Although, despite extending first-rate marketing attempts, there have been increasing scenarios where it’s no longer possible to observe whether a conversion has occurred, especially for view-through and cross-device conversions. From cookie restrictions in-browser to holes left from cross-device shopping, marketers often have to make sense of a frequently incoherent view of consumer behaviours. Even on a single device, different browsers can persuade different paths to purchase. This is where Attribution Models come into play. As defined by Google;
"An attribution model is the rule, or set of rules, that determines how credit for sales and conversions is assigned to touchpoints (impressions and clicks) in conversion paths"
How does Conversion Modeling work?
Google Ads models website conversions that should be ascribed to clicks and views but may go unobserved. This is done by recognizing divisions of traffic with observable conversions and applying statistical techniques to shares of traffic where data is partial or missing entirely. When unable to observe a subset of conversions, the conversion modeling tool uses machine learning to quantify the impact of set marketing efforts. Leveraging visible signals such as date and time, device type, and conversion type, the tool produces an authentic, aggregate view of user behaviour. These unobserved events are then incorporated into your account's reports, along with the noted conversions. To secure accurate conversion modeling for your account, consider leveraging Google's global site tag or implement Google Tag Manager to record as many authentic conversions as possible. This will help the conversion modeling tool piece together probable actions and apply its unique set of statistical modeling techniques.
In Campaign Manager 360, the Attribution Modeling Tool is again used to compare different forms of assigning "credit" to particular sources of traffic. On September 28, 2020, to better track missed conversions, all attribution models in both Display & Video 360 and Campaign Manager were opted into conversion modelling by default. The tool provides the following common attribution models:
- Floodlight Model: Attributes the conversion to the last click made by the user before purchasing. If there was no click, the model attributes the value to the last impression.
- Last Interaction: Conversion value is attributed to the last channel with which the customer interacted before buying or converting.
- First Interaction: Conversion value is attributed to the first channel with which the customer interacted.
- Linear Model: Equal credit is given to each channel interaction that ultimately resulted in a conversion.
- Time Decay: A majority of credit is given to the touchpoints that are nearest to the time of conversion. This model has a half-life of 7 days, meaning that a touchpoint 7 days before a conversion will get half the credit of a touchpoint on the same day as the conversion or sale.
- Position-Based Model: Credit is split between the first and last interactions made resulting in a conversion.
- Social Model: Based on the linear model but impressions are weighted to account for social interactions. The default weightings are:
- Impressions without social engagements = x0.5
- Impressions with any low-value social engagements but no high-value (Has engaged, but not shared) = x0.75
- Impressions with any high-value social engagements (extend reach to other users) = x1.5
Various factors, such as business goals or buying cycles, can make one model more suitable than others. When Google is unable to directly observe the events leading up to a conversion, it relies on aggregated and anonymized data to estimate the unnoticed touchpoints. You can get a better idea of the return on investment (ROI) for each marketing channel by analyzing each attribution model. Each model has its own unique set of strengths and weaknesses. For example, if your site doesn't have a quick conversion cycle, using the default “Last Interaction” method may not be the best choice for your campaign. On the other hand, if your focus is on learning where your most recent customers came from, leveraging “Last Interaction” might be worth considering, however, this ignores the other ad interactions customers may have had along the way.
In addition to the above seven common attribution models, Google offers an additional model called the Data-driven attribution model which recognizes conversion data to determine the actual aiding of each ad engagement across the conversion path. This particular attribution model specifically compares the paths of customers who did convert to the paths of those who did not. The model then identifies patterns amongst those ad interactions that resulted in a conversion and helps drive additional conversions at the same CPA to help advertisers identify which ads have the greatest effect on their business goals. The Data-driven attribution model works by pulling data from your account to determine which ads, keywords, and campaigns have the greatest impact on your goals and assigns credit for conversions depending on how people engage with the ads.
The attribution model you choose only affects the conversion action to which it's applied. You can set the attribution model when you're setting up your conversion action, or change the attribution model for an existing conversion action, so consider replacing your attribution models from "Last click" to "Data-driven" to help determine which ad interactions are most impactful.
Overall, attribution models give you more control over how much credit each ad interaction receives for your conversions. Understanding how consumers engage with your content and behave across the web can lead to significant insights that can make your ad inventory more relevant, your campaigns more compelling, and improve overall conversion rates. Whichever model you choose, consider testing your theories through experimentation. By testing new strategies and evolving your marketing practices, you can sustain and advance your revenue — while respecting people’s preferences for privacy.
Now that you have a better understanding of the importance of conversion modeling, you may want to take the time to review your existing models and update your preferred modeled conversion settings. Still not sure which model will help you meet your campaign goals? Reach out to the Full-Service Programmatic experts at Hotspex Media today to start measuring the metrics that matter most.