Learn About Data Cloud Insights

Learning Objectives

After completing this unit, you’ll be able to:

Explain how insights are used in Data Cloud.
Describe the difference between calculated and streaming insights.

Data Platform Insights

Salesforce Data Cloud (formerly known as Customer Data Platform) brings together data from across your company to drive personalization, engagement, and provide a single view of your customer. You can also enrich this data using insights, which are complex metrics you can define and calculate using SQL (Structured Query Language). In this module, we cover the types of insights you can create and how to use these insights to optimize your data in Data Cloud.

Types of Insights

First, there are two types of insights: calculated and streaming.

Calculated insights are used to query and create complex calculations based on stored data.

Streaming insights are queries based on real-time data.

Let’s review a few more differences between the two.

	Calculated Insight	Streaming Insights
How is data processed and collected?	Data is processed together as a unit in high-volume batches.	Data is processed from streaming data sources such as from the Web or Mobile SDKs.
What types of data are used?	Calculated insights can be created from any data.	Streaming insights can only be created from engagement data, not streaming profile data.
How can you use these insights?	Calculated insights can be used to define segment criteria and personalization attributes for activation using metrics, dimensions, and filters.	Streaming insights help build time series aggregation in near real time that can be used to drive orchestration or data actions.
How can these insights be shared?	Calculated insights are packageable and can be shared with other instances of Data Cloud.	Streaming insights can be mapped to different objects from Web and Mobile SDK and Marketing Cloud Engagement data streams.

Now let’s review some use cases for each type of insight.

Use Calculated Insights

Data Cloud uses direct and related attributes to build segments. Within the segmentation interface, you can use operators to do simple aggregations to filter your data. For example, you can segment based on an attribute (birth date), an operator (is before), and a value (date).

Segmentation rule for birth date before December 1.

However, there are times when segmentation operators don’t fully meet your needs. Let’s say you want to find customers who have an average order spend of $50 or more. For max, min, and sum operators, you need to use calculated insights. Once an insight is created, you can then use it to narrow your segmentation.

Segmentation using a Calculated Insight for average order spend.

Here are a few more metrics that can be created with calculated insights.

At least 5 email views per quarter
Cart value over $500
Customer rank greater than 3
Total sales amount greater than $1000
Open tickets greater than 1 this past year

You can also use calculated insights to simply clean data before you segment—such as formatting data by rounding. Now let’s take a look at streaming insights.

Use Streaming Insights

While calculated insights focus on data as a whole, streaming insights focus on data at a specific time. There are many ways streaming insights can be used, but most involve taking action on data notifications. Let’s review a few industry use cases where creating a streaming insight would be helpful.

Industry

Use Case

Retail

Challenge: A Northern Trail Outfitters (NTO) customer enters a retail store in Seattle. Isabelle, an NTO marketer, wants to send push notifications with a coupon to any customer entering a store for the first time in 3 months.

Solution: NTO creates a streaming insight based on real-time customer geofence data. When a customer enters a store’s geofence and hasn’t appeared in the data for at least 3 months, a workflow is initiated. The workflow adds the customer to a reengagement journey in Marketing Cloud Engagement.

Hospitality

Challenge: Gallagher Resorts sales reps want to know which of their customers have the highest points balance, especially when they call their top customers at year end.

Solution: Gallagher Resorts creates a streaming insight that automatically updates a Salesforce report with the top 50 customers based on points balance. Sales reps have a dashboard created on their homepage that highlights which of the top customers are in their portfolio. This allows them to prioritize a phone call with that customer.

Healthcare

Challenge: In order to streamline an agent’s conversation, Bloomington Caregivers wants to know what troubleshooting steps a customer has taken before they call support.

Solution: Bloomington Caregivers sets up a streaming insight to track customers visiting their top knowledge articles. Then, a case is automatically created for any customer visiting the page more than 5 times or with an average viewing time of over 10 minutes. Support agents can proactively reach out to those customers or have a case already created if a customer calls support directly.

Define Clear Goals for Your Insights

Regardless of the type of insights you want to use, you should be aligned with your team on the outputs and goals they expect from the insights. Work with your team to predefine these use cases and identify where the data is stored before you build them in Data Cloud.

Next Up: Create Insights

Now you know the types of insights available and common ways to use them. In the next unit, we cover the methods for creating both types of insights in Data Cloud.