Predictive Segmentation with Einstein in Salesforce Marketing Cloud

Einstein Engagement Scoring

Einstein will help you understand your customers better and help you to visualise and predict their behaviour. In other words, it’s a powerful tool that uses artificial intelligence (AI) to provide advanced and predictive segmentation. Inside Einstein you will find Engagement Scoring, a data science driven feature that assigns scores for each subscriber. It analyses the data from all of your past emails campaigns to segment the subscriber based on their engagement. In short, it will predictive the customers engagement and target the optimal audiences for your marketing offers.

Einstein Engagement Scoring will begin with analysing your subscriber base and and predict their engagement across four different behaviours. It analyses how likely they will open the email, engage with it, unsubscribe from the emailing list or convert. The predictions are made for the next 14 days and the data is refreshed daily. Moreover, Einstein will tell you how healthy or engaged your subscriber base is compared to the industry’s benchmarks. To understand the factors behind the predictions, go the the prediction detail page. Here, you will see top positive and top negative predictors from the prediction model.

To get the most out of Einstein Engagement Scoring, start to create audiences in Audience Builder. Since each score is stored in a Data Extension, you can easily put them to use and create audiences. Find the engagement scores predefined in the Audience Builder for easy use. Use these to filter out audiences, like at risk subscribers, or segment your audience for personalisation. When you ready to send an email, just choose the audience filtered by the engagement score. In that way you can send targeted messages to differet subsribers and maximize the engagament.

Additionally, you can use Einstein together with your customer journeys in Journey Builder. For example, use the score as an entry event or use it to test different channels. For instance, if someone is not likely to open your email, you can show them an add or send them a sms message. Above that, you can also query the engagement scores using SQL.

Now, what about the Predictive Segmentation? Inside the tool you will find segments with audience personas in four categories:

• Loyalists – Subscribers who are most likely to open your emails and engage with them, clicking on the content.
• Window shoppers – Subscribers who are likely to open your emails but are unlikely to click on the content.
• Selective shoppers – Subscribers who are unlikely to open your emails but when they do open, are very likely to click on the links.
• Win-back or Dormant – Subscribers who are unlikely to open or click on your emails and do not engage with them.

This predictive segmentation will help you filter your audience, personalize content in your emails or optimise the use of your channels. You can start by setting individual goals for each of the segments and optimize the message based on what suits best with their behaviour. Summarising, with Einstein Engagement Scoring you will save time, improve the ROI of your marketing automation, and elevate your email marketing efforts.

Customer Success Story – Room & Board

Room & Board, an interior company, are using predictive intelligence to power their recommendation. Since they started to use recommendations in the email campaigns, customers who engaged with those emails converted 40% more.

Einstein Marketing Insights

Get deeper into Einstein’s head and learn more about Einstein Marketing Insights and how you can uncover hidden optimisation opportunities.