Microsoft Dynamics 365 for Customer Insights (DCI) SaaS offering provides big data analytics and insights for sales, marketing and service functions. To extract the most out of DCI requires bringing in data from different sources. One of the most important data sources is Dynamics 365 Customer Engagement. In this article, I want to walk you through the process of managing data sources and integration with DCI. The process of extracting insights is a typical walk through steps of scoping insights requirements, modeling data, implementing data map, data preparation, transformation and flow, and finally, data ingestion to DCI.
The journey of DCI analytics starts with defining the insights and a use case that fit the insights. Take the example of customer 360 view with DCI for a retail banking customer service center.
Bank client, Mike Price, just enters one of the bank’s branches. The bank has his demographic profile in DCI. DCI shares the client’s phone number and shows the visuals and pictures of Mike to identify him when he is sitting in front of me. It pulls up rich set of various KPIs calculated inside customer insights by its powerful KPI processing engine. The bank want to see and understand who this customer is. Based on all the raw data, DCI constantly calculate those measures. While it spews out insights, DCI could also share trivial measures, like Mike’s checking account balance, savings account balance, KPIs of KPIs, and so on. DCI provides widgets that highlights Mike’s communication preferences, what channels Mike is using and how Mike is interacting with the bank. DCI shows the financial instrument, and last but not the least, it shows end-to-end customer journey with all interactions and transactional data that you might want to look at.
Building Retail 360 experience requires multiple steps on data management side. It starts with identifying the data sources and corresponding connectors to bring data in from those sources, modeling the data, transforming the data, and then activating the connector to start the complete end-to-end flow.
The key to DCI success is in its ability to bring in data from multiple sources, and continuously extract insights from data. Once the scope of the use case is defined (e.g., retail bank customer 360 view defined above), the most important task is to bring the data in from multiple sources and model it in DCI. DCI excels in integrating data from multiple different kinds of data sources characterized by different attributes.
A data connector mechanism facilitates data source connection and data pull to DCI. Connector performs following functions, including
Data source’s attributes compound connector’s complexity. Data sources differ based on multiple factors, including, but not limited to the following:-
Many other factors contribute to the complexity of connectors for DCI, e.g., data access interface, data model – including the structure or lack thereof, of schema, operations support, bulk export capabilities, governance requirements, interaction model, speed and volume of data change in the source, etc.
To build a retail bank customer 360 with demographics data from all other sources starts with identifying data sources that meet the use case needs. It could be data from CRM systems, social media, web logs, mobile application logs, or any other database.
There are three different ways to bring data into DCI. These are:-
Ingestion mechanism choice depends upon multiple factors, most important being, and availability of custom connector for your data source or not.
DCI data model models data after schema.org concepts. Key DCI data building blocks are profiles, interactions, and links.
Depending upon the type of connector, DCI may provide an out of the box data model. DCI already does that for Dynamics 365 Customer Engagement connector. If not, users need to create their own data model or enhance an existing one when bringing in data from a new source.
Today, there are two out of the box connectors available in DCI December public preview version, Microsoft Dynamics 365 Customer Engagement connector and Azure Blob connector. While Dynamics 365 Customer Engagement connector provides very structured connectivity to a clear single data source, Azure Blob connector provides free form data ingestion mechanism to scale amount of data and variety of data ingested in DCI.
Key differences between Dynamics 365 Customer Engagement connector and Azure Blob connector are captured in the table below.
Users can build their own connectors for DCI to emulate some of the custom connector functionality. DCI models data after schema.org concepts. DCI users can build the data model corresponding to the source to comply with DCI data model. Next, they can combine the Azure Blob connector and point ingestion mechanism to create the connector with custom orchestration, controls and policies. In particular, users will need to implement the following key functionality to create a custom connector.
Additionally users need to define and implement other aspects of connector like data movement orchestration, sync policy, frequency of data updates, speed of data copy, resources auto-scaling, error handling, etc. for a seamless connector experience.
Once a data source connector is added to DCI, activate the connector to let the data flow in to begin the process of insights creation. Exact technical details and steps of how to add connector data source and deploy it can be found in DCI Step-by-step configuration guide.
– Dev Vidhani