Attribution Webinar Recap: Making Attribution work for Your Business
November 15th, 2012 | Published in Google Analytics
On Friday, November 2, following our public whitelist of the Attribution Modeling Tool, Bill Kee (Product Manager, Google Analytics) and Neil Hoyne (Global Program Manager, Attribution), came together to lead the 5th and final webinar in our series on marketing attribution. They identified opportunities in the customer’s journey from introduction to conversion, including:
You can also catch up with our entire attribution webinar series, which included:
As has been our tradition throughout this webinar series, we’d also like to provide responses to some of the most common and most interesting questions we received during the webinar.
Questions
What business variables influence the decision on an Attribution Model?
Any factor that could influence your business or marketing efforts, including weather, pricing and competitive behavior, could have an impact your attribution decisions. Still, we suggest that advertisers focus on those efforts that could have the largest effect on their business, usually by conversion volume as well as those that they can more easily control (paid search vs. organic search or direct traffic) for the basis of experimentation.
How is the social engagement metric calculated?
Social engagement is measured any time a user clicks from a known social network, such as Facebook, Twitter, Google+ or over 400 others, to the advertiser’s website. At this time, no interactions that occur within the networks themselves, such as a “like” are presented within the Attribution Modeling Tool.
Could you further elaborate on how conversion paths are presented when a user converts multiple times within the 30-day lookback window?
Each conversion has a unique path, which includes all of the interactions the converting user had in the 30 days leading up to the conversion. When the same user converts multiple times, the conversions are treated separately. For example, is a user clicked through from Display, and completed conversion #1, this conversion would have a path length of one from the channel “Display.” If the same user subsequently clicked through from Paid Search, and completed conversion #2, assuming the original Display interaction occurred within 30 days prior to conversion #2, a second conversion path would be recorded with a path length of two: Display, followed by Paid Search.
If we submitted our account to the Attribution Modeling Tool whitelist, how long will it take until we begin to see this feature available in our Google Analytics account?
We understand how important attribution is to your business, and are incredibly grateful for all of the interest that has been shown in the modeling tool since the announcement of the public whitelist. As such, we are working as quickly as we can to add new customers to the tool and will continue to post any available updates directly on the signup form. Once your account has been whitelisted, you’ll see the Attribution Modeling Tool listed within the Multi-Channel Funnels reports, under Conversions.
Could you provide step-by-step details on how to build the models Bill described during the webinar?
We created two custom models to show examples of the types of weighting you can apply using the model builder. The first model, called “Upper Funnel” emphasizes interactions earlier in the path, from channels that are focused on introducing and informing customers, and discounts channels that may be more navigational, like branded search. The second model, called “Lower Funnel” gives more weight to marketing interactions at the end of the conversion path, but does not solely give credit to the last interaction, and excludes direct interactions that are last in the path, giving credit instead to other marketing touch points toward the end. By comparing both models to the Last Interaction model, you’re able to see the contrasts in credit given to channels, and see whether marketing efforts play the roles you think they do or not.
Here are the rules for the “Upper Funnel” model.
Upper Funnel Model, step 1: Click on the model selector then “create new custom model” to open the custom model builder, and enter details as pictured (click to enlarge the image):
Upper Funnel Model, step 2: Turn on “apply custom credit rules” in the custom model builder, then enter model details as pictured (click to enlarge the image):
And here are rules for the "Lower Funnel" model.
Lower Funnel Model, step 1: Click on the model selector then “create new custom model” to open the custom model builder, and enter details as pictured (click to enlarge the image):
Lower Funnel Model, step 2: Turn on “apply custom credit rules” in the custom model builder, then enter model details as pictured (click to enlarge the image):
Marketing attribution is a challenging yet worthwhile pursuit. Our hope is that this webinar series will help you as you begin (or continue) your attribution journey. For more information on the Attribution Modeling Tool, please visit our website and the Google Analytics help center.
Happy analyzing!
Sara Jablon Moked, Product Marketing Manager for Conversion and Attribution
- Google’s recommendations for how companies should structure their own attribution programs.
- Basics on the methodology and configuration of the Attribution Modeling Tool, and how to create custom models that can improve your business’ performance.
- Identifying specific opportunities in attribution from brand-to-generic trends to position-based weighting.
You can also catch up with our entire attribution webinar series, which included:
- an overview of our research on how the industry approaches attribution (watch here),
- the foundational steps for attribution using Google’s tools (watch here),
- intra-channel attribution with Search Funnels in Google AdWords (watch here),
- cross-channel measurement with Multi-Channel Funnels (watch here),
- and finally, our most recent webinar on strategies for the Attribution Modeling Tool (watch here).
As has been our tradition throughout this webinar series, we’d also like to provide responses to some of the most common and most interesting questions we received during the webinar.
Questions
What business variables influence the decision on an Attribution Model?
Any factor that could influence your business or marketing efforts, including weather, pricing and competitive behavior, could have an impact your attribution decisions. Still, we suggest that advertisers focus on those efforts that could have the largest effect on their business, usually by conversion volume as well as those that they can more easily control (paid search vs. organic search or direct traffic) for the basis of experimentation.
How is the social engagement metric calculated?
Social engagement is measured any time a user clicks from a known social network, such as Facebook, Twitter, Google+ or over 400 others, to the advertiser’s website. At this time, no interactions that occur within the networks themselves, such as a “like” are presented within the Attribution Modeling Tool.
Could you further elaborate on how conversion paths are presented when a user converts multiple times within the 30-day lookback window?
Each conversion has a unique path, which includes all of the interactions the converting user had in the 30 days leading up to the conversion. When the same user converts multiple times, the conversions are treated separately. For example, is a user clicked through from Display, and completed conversion #1, this conversion would have a path length of one from the channel “Display.” If the same user subsequently clicked through from Paid Search, and completed conversion #2, assuming the original Display interaction occurred within 30 days prior to conversion #2, a second conversion path would be recorded with a path length of two: Display, followed by Paid Search.
If we submitted our account to the Attribution Modeling Tool whitelist, how long will it take until we begin to see this feature available in our Google Analytics account?
We understand how important attribution is to your business, and are incredibly grateful for all of the interest that has been shown in the modeling tool since the announcement of the public whitelist. As such, we are working as quickly as we can to add new customers to the tool and will continue to post any available updates directly on the signup form. Once your account has been whitelisted, you’ll see the Attribution Modeling Tool listed within the Multi-Channel Funnels reports, under Conversions.
Could you provide step-by-step details on how to build the models Bill described during the webinar?
We created two custom models to show examples of the types of weighting you can apply using the model builder. The first model, called “Upper Funnel” emphasizes interactions earlier in the path, from channels that are focused on introducing and informing customers, and discounts channels that may be more navigational, like branded search. The second model, called “Lower Funnel” gives more weight to marketing interactions at the end of the conversion path, but does not solely give credit to the last interaction, and excludes direct interactions that are last in the path, giving credit instead to other marketing touch points toward the end. By comparing both models to the Last Interaction model, you’re able to see the contrasts in credit given to channels, and see whether marketing efforts play the roles you think they do or not.
Here are the rules for the “Upper Funnel” model.
Upper Funnel Model, step 1: Click on the model selector then “create new custom model” to open the custom model builder, and enter details as pictured (click to enlarge the image):
Upper Funnel Model, step 2: Turn on “apply custom credit rules” in the custom model builder, then enter model details as pictured (click to enlarge the image):
And here are rules for the "Lower Funnel" model.
Lower Funnel Model, step 1: Click on the model selector then “create new custom model” to open the custom model builder, and enter details as pictured (click to enlarge the image):
Lower Funnel Model, step 2: Turn on “apply custom credit rules” in the custom model builder, then enter model details as pictured (click to enlarge the image):
Marketing attribution is a challenging yet worthwhile pursuit. Our hope is that this webinar series will help you as you begin (or continue) your attribution journey. For more information on the Attribution Modeling Tool, please visit our website and the Google Analytics help center.
Happy analyzing!
Sara Jablon Moked, Product Marketing Manager for Conversion and Attribution