Get suggested segments
Up until now, we have been building segments manually based on criteria we define as we build the segment. Often, that can only get you so far. Many times, organizations can miss opportunities to target the right customers because they aren't able to easily identify who they should be targeting. As your customers evolve, you might find that there are emerging customers that you should be focusing on, that weren't that obvious initially.
Customer Insights - Data helps you discover interesting segments of your customers with the help of an Artificial Intelligence (AI) model. This machine learning powered feature can suggest segments based on measures or customer attributes. It can help improve your overall KPIs and help you to better understand the influence of attributes in context of other attributes.
For example, you can use segment suggestions to identify if customers in specific segments have a favorite time of day they like to engage. With Segment suggestions, you can identify which customers prefer a specific time of day and have the system help you create segments based on that data. Customer Insights - Data lets you configure suggested segment capabilities based either measures or activities.
Generate suggested segments
Measure segment suggestions can be created by navigating to Segments and Selecting the Suggestions (preview) tab. When you select Get new suggestions, it starts with a guided experience.
You can base segment suggestions on either:
A Measure
Or a Customer attribute
You also need to define which attributes to consider as influencing attributes. Based on your needs you can select a single attribute or multiple attributes. Selecting multiple attributes with improve the chances of evaluating how they influence the primary attribute.
Suggested segments based on measures
As a user of Customer Insights - Data, you likely have a series of measures created that help track your Key Performance Indicators (KPIs). It's important to understand how certain attributes influence this KPI to help you create better segments and run more highly targeted campaigns. For example, let's say that you track a measure called TotalSpendPerCustomer. As a business, you want to see this number grow.
When you create a segment suggestion that is based on a measure, you can select which attributes you want to assess for influence. Maybe you say membership tier, membership period, and occupation.
With this information, Customer Insights - Data can then suggest a segment that tells you who is the biggest influence of that measure. For example, it might tell you that Accountants who are Gold members, and who have been with your business for at least five years are the biggest influence of TotalSpendPerCustomer.
Suggested segments based on customer attributes
Instead of a measure, you can choose a customer attribute instead. Based on your choice of influencing attributes, the AI model creates a series of suggestions that show how the selected attributes influence the primary attribute. For example, Let's say that you choose Rewards Member (Yes/No) as the primary attribute. You then use Tenure, Occupation, and Number of Support Tickets are set as other influencing attributes. The AI model could suggest segments indicating mostly IT professionals with tenure over two years are rewards members. Another suggestion could highlight that accountant with tenure over one year and fewer than three support tickets are rewards members.
When you select a measure or customer attribute for the suggestion, once the AI model generates the suggestions, you can find them listed on Segments > Suggestions (preview). You can select a suggested segment to review the details of that suggestion. You can also review the attribute values or rules that the AI model learned to suggest the selected segment.
If you want to save the suggestion as a segment, you can select the segment you want to save and choose Save as segment. After you save segment, it shows in the list of segments on the All-segments tab. It can now be refreshed, edited, or deleted like any other segment.
Suggested segments based on activity
Besides discovering segments based on Measures you define, you can also discover interesting segments based on customer activity data that is ingested to Customer Insights - Data. Examples of activity data are transactions, support call duration, purchases, or returns. To suggest segments, activity data gets analyzed for recency, frequency, and monetary value (or duration). Instead, you can generate suggested segments to improve a measure or better understand what influences an attribute.
With activity data available in Customer Insights - Data, we can generate suggestions that represent customer groups:
Most active customers
Customers that made the most purchases
Customers that generated the most revenue
Customers who weren't active lately
Customers who frequently interact with your business
If you have a retail business, you could find out which customers generate the most revenue and reward them with a coupon. Or you can identify occasional customers and offer them to join a rewards program, so they visit your business more often. If you're in the healthcare business providing public healthcare and your goal is to minimize the expenses for individual patients. A way to do so could be to reduce recurring visits by providing the best possible care in as few visits as possible. In this case, your goal is to keep the visit frequency low and minimize recurring cost for the patients. Or you can identify segments of patients who have frequent appointments and high recurring costs and analyze these cases to improve the treatment of the individual.
Suggestions are generated based on the selected input data.
Customer profiles: All customers or members of a specific segment.
Time period: Last month, last year, or any custom time frame.
Activity type: purchases, retail transactions, online transactions, customer support cases, subscriptions, and so on.
Table in Customer Insights - Data that contains the activity data: The Unified Activity table or the table for a specific activity.
Dimensions to include: Recency, frequency, or monetary dimension, depending on your business requirements.
Activity segment suggestions can be created by navigating to Segments, selecting Find New Suggestions, and selecting See or anticipate customer behavior. When configuring activity suggestions, you need to set the following options:
Customers: Defines the customers that you want included. You can select all customers choose from a specific segment.
Activity: Specify the Activity Type that you want to use along with the tables that describe the activity.
Preferences: Define the time-period to consider, factors for suggestions, and then map the attributes.
Once the AI model generates the suggestions, you can find them listed on Segments > Suggestions (preview) in the Activity-based suggestions section. Select a suggested segment to review the details of that suggestion. You can also review the attribute values or rules that the AI model learned to suggest the selected segment.
If you want to save the suggestion as a segment, you can select the segment you want to save and in the side pane select Save as segment. After you save the segment, it shows in the list of segments on the All-segments tab. It can now be refreshed, edited, or deleted like any other segment.
Select See suggestion on a suggested segment to view the details of that segment. The side pane provides details like the extent of each dimension in comparison to the target group. It also highlights the number of potential members in the segment and the corresponding percentage of the total customers. If you want to keep the suggestion as a segment, select Create segment.