Chart of the Day: Use Machine Learning AI to understand the propensity of your site visitors to convert
I'm reviewing a different type of Chart today since while reviewing Google's Demo Account while working on some new learning materials for our members I noticed the intriguing new Conversion Probability feature Beta.
Here's what it shows for the one hundred thousand odd sessions in the Demo account:
You can see that it breaks down these sessions into likelihood to convert, so if we can learn more about what turns the high-converting visitors onto our products, we have more chance of finding and converting more visits in future.
As we explain in our AI and machine learning guide, for us, the most exciting marketing application of artificial intelligence is using machine learning to learn from historic interactions with our audiences see what influences their propensity to convert. Using this insight we can tailor our communications to be more relevant.
Here, Google analyses historic visits of sites which have at least 1,000 e-commerce transactions to see which of all the variables available in analytics like visitor source, content consumed and path determine propensity to purchase. You can use the conversion probability metrics to segment in Google Analytics so that if you choose a higher probability segment such as 21-50 or 51-100, you can assess :
- Which channels, keywords, and campaigns deliver highly engaged users.
- Which conversion paths are most effective, and where along the path can you deliver the most effective advertising.
- Create remarketing audiences based on your users who are more likely to convert and publishing those audiences to your various marketing platforms like AdWords and DoubleClick Bid Manager.
- Publish these audiences to Optimize so that you can understand exactly which refinements to your site content deliver the highest likelihood of conversion.
Hopefully, this shows the potential power of this exciting technique to retailers and other ecommerce businesses that have Ecomm transactions set up in Google Analytics.