Last year, at Signals 19, we announced our big news — the arrival of our Customer Engagement Suite (CES) and the foundational Engagement Data Platform (EDP), which provides a unified data, decisioning, and delivery engine for all of our offerings. Since our Signals launch, we have been collaborating with customers who have been critical in helping drive our roadmap.
Just this week, we completed the latest release of the EDP, which included 150+ enhancements and optimizations, as well as achieved key data ingestion milestones, hitting numbers such as 50M-250M email events/hour and 5-10M person profile updates/hour. In addition to those milestones, we’ve added new updates that you can check out below.
Machine Learning (ML): Send Time Optimization (STO) and Propensity Scoring
Almost everyone I have spoken with about the platform in the last year is interested in leveraging the EDP for Machine Learning (ML). One of the many benefits of an EDP for the marketer is having a vast amount of useful data in one place for decisioning and delivering personalized marketing campaigns.
ML accelerates insights and decision making, which helps the marketer drive better targeting, personalization, and engagement. An example of that is Send Time Optimization (STO), which uses ML to determine the best time to send an individual an email, resulting in higher open and click-through rates, as well as increasing customer engagement. The benefit to the consumer is that emails are less intrusive and more meaningful. Another example of ML in the EDP is Propensity Scoring, which helps marketers determine how likely a customer is to engage in a predictive behavior. Examples include identifying customers with a propensity to open or click, identifying those predicted to engage less frequently, or even opt-out.
Both STO and Propensity Scoring help marketers quickly optimize engagement strategies that bring increased marketing results.
Given the very nature of the EDP as a unified data hub, decisioning and delivery engine, marketers need access to dashboards which surface program performance and customer behavior. We’ve created new dashboards in the EDP that are designed to transform data into useful and actionable insights. An example is the ability to filter by “organizational units,” as well as loyalty dashboards.
The new reports and dashboards in the EDP enable marketers to explore data and curate custom content that can be saved and distributed to others in the organization.
Core to the EDP value proposition is the ability to integrate with other marketing applications in customer environments, making it a centralized marketing brain. It has a robust set of integration capabilities including batch, real-time, inbound and outbound. We have been continuously enhancing these capabilities including an improved user interface, simplified formats for activities and orders, more object types in import and exports, including exports based on an analytics query, and the ability to syndicate disposition data to a customer’s data warehouse.
There are many more important features, including the first phase of Multi-Business Unit support, an enhanced Events API, support for non-unique email addresses, and the ability to trigger actions on selected disposition events such as bounces and unsubscribes.
Key areas of investment for the EDP
Enhanced Machine Learning — Deliver additional machine learning models, such as Propensity and Clustering, and improvements to the ML framework, such as the ability for citizen data scientists to configure and tweak out-of-the-box models.
EDP Sending via Marigold Engage+ — Support features that existing Marketing Suite customers commonly use, such as looping blocks and A/B/n testing, when sending from EDP. Also, surfacing EDP and Loyalty data and objects better in Marigold Engage+.
World-Class Performance — We will continue to improve performance, especially for segmentation and analytics, with add-ons such as an in-memory segmentation engine and a built-in aggregation layer.
If you have any questions about our EDP updates or anything else, please reach out to us.