For specific service scenarios, marketing personnel analyze and evaluate customer segments provided by the Interactive Customer Segmentation feature, and then create value customer segments for the service scenarios, improving the marketing precision.
Application scenario of the Lookalike Analysis feature
Application scenario for Lookalike analysis
A carrier plans to retain customers predicted to churn. The carrier uses the Upgrade package to send phone segment as the seed customer segment, and selects top 10% customers who are similar to customers in the seed customer segment from the Forecast offline segment as the target customer segment.
Customers covered by a tag or customer segment can be used as the seed customer segment and target customer segment. The involved thresholds are configurable. Figure 6-43 shows the Lookalike customer segment creation page. The requirements on the seed and prediction customer segments are as follows:
-The number of customers in the seed customer segment must be between 500 and 100,000.
-The number of customers in the prediction customer segment must be between 0 and 10,000,000
To achieve good analysis effect, the number of customers in the seed customer segment is at least 500. To ensure system performance, the maximum number of customers in a customer segment is limited to 10,000,000.
Lookalike customer segment creation
- Lookalike Analysis
- This function extracts public features of customers in the seed customer segment, and then selects customers whose features are similar to seed customers from the prediction customer segmentCustomers covered by a tag or customer segment can be used as the seed customer segment and target customer segment.Customer segments can be created in Lookalike mode.