Salesforce Marketing Cloud Einstein Engagement Scoring
Our second topic in the “Einstein Components in Salesforce Marketing Cloud” series is Einstein Engagement Scoring (EES). Once again, it's about optimizing the engagement of your own recipients, but in a slightly different way than Einstein Engagement Frequency (EEF). While with EDF, the optimal quantity The number of emails to be sent per contact is determined, EES offers a dashboard from which the marketer gains insights into the reaction its recipients can gain access to the emails sent. This in turn should serve to better tailor communication to the individual contact. How does that work?

The Einstein Engagement Scoring Dashboard consists of five different elements. The Email Engagement Prediction tile divides recipients into four different quadrants in a heat map: “Loyalists” (high open and click rate), “Window Shoppers” (high open rate, low click rate), “Selective Subscribers” (low open rate, high click rate) and “win-back/ dormant” (low open and click rate). The heat map helps the marketer understand what the percentage distribution between these four personas looks like. Based on the personas, segments can also be created in Audience Builder, which are then played with appropriate communication. For example, Dormant Subscribers can receive a reactivation campaign or get Selective Subscribers to open emails more often with content that has been proven to match their interests. The Subscriber Retention Prediction, Email Open Prediction, Email Click Prediction, and Web Conversion Prediction tiles each represent the probability that a contact will remain signed in, open an email, click in an email, or convert in the next 14 days. For the latter, the Collect Tracking Code must of course be integrated on the target page so that a conversion is recorded. The calculated probability is given as a percentage and rated on a four-level scale (from “poor” to “excellent”). When classified into these four levels, the company's own results are compared with standard values from the digital marketing industry. Each tile also provides an assessment of the “model confidence”, i.e. the trustworthiness of the calculated results, which depends on how comprehensive the collected data is.
In summary, the EES dashboard can therefore provide a very good overview of the engagement of its own recipients — subject to a few caveats: First of all, the dashboard only works if (almost) all recipients are tracked, otherwise the figures are severely distorted and therefore no longer usable. In addition, segments based on the personas can only be created in Audience Builder, which requires licensing of this product. And last but not least, classifying your own results on a scale based on cross-sector averages is only of limited significance. Our conclusion: The EES dashboard offers interesting insights into the behavior of your own recipients when certain requirements are met and you know how to interpret the results correctly.
Here you can find the Einstein Engagement Scoring Documentation.