View Entry

Campaign Response Management 

Wed November 20, 2019 01:05 PM

Business Value and Use Case

- Solution Story

This Industry specific solution provides a Campaign response model to help companies in selecting high-quality receptive customers for their future campaigns. It will enable the companies to invest in direct marketing with an effective and different selection of contacts.

How are we doing that?

By analyzing the customer’s response to a marketing campaign based on their past characteristics and behavior (e.g. Customer Demographics), we build a predictive model that helps us in predicting the probability of prospective customers responding to the new campaign.

This solution is built using the IBM Cloud Pak for Data (CP4D) platform which supports:

  1. Integration of all the data from enterprise RDBMS, multiple cloud platform applications from where historical Campaigns were run and the response prediction models from R Studio and Python notebooks
  2. Faster decision making and creation of new digital products and services – seamlessly we were able to collate the necessary datasets for modeling, import and tweak existing models and deploy them as API and also push the prediction output to embedded Dynamic dashboards all in one single platform all within 2 days of getting access to the secure and enterprise governed IBM Data & AI platform instance
  3. Efficiency and cost savings instead of loading the customer with an annoying direct marketing outreach program. This indirectly helps in increasing NPA scores of existing customers who positively respond to the marketing campaigns.

The prediction algorithm captures the dependencies of customer attributes on their response, and its effectiveness is quantized and plotted using Gain and Lift charts.

The visual dashboards help us reaching the target within the marketing budget. Depending on the costs associated with sending direct mail and the expected revenue from each responder, this campaign response model can be used to decide upon the optimum number of customers and their respective segments to contact. It would help us in getting a tipping point at which we have reached a sufficiently high proportion of responders, and where the costs of contacting a greater proportion of customers are too great given the diminishing returns.






Demo Scenario

CUSTOMER DEMOGRAPHICS

This section talks about the Exploratory Data Analysis (EDA) of customer attributes that gives the information of respondents in percentage by Age Group, Marital Status, Occupation, and Education.

HISTORICAL CAMPAIGN PERFORMANCE

For the same customer segment, here we’ll analyze the historical campaign performance. This captures the attributes like contact duration, contact month, campaign channel, number of previous attempts to acquire the customers.

Based on this, we optimize the strategy for non-respondents.

POTENTIAL TARGET CUSTOMERS

The model buckets the customers based on their receptivity towards the campaign into 3 segments i.e. highly receptive, medium receptive and low receptive customers.

The receptiveness is calculated from the probability of a customer likely to respond to a campaign.

The probability/score from the model is ordered in the decreasing order to get the top customers likely to respond.

MODEL EVALUATION

This tab talks about the effectiveness of the predictive modeling technique that returns the percentage of customers who have responded (development) /will respond (validation) which is analyzed through plotting Gain chart and lift curve.

Gain chart: Gain at a given decile level is the ratio of the cumulative number of targets up to that decile to the total number of targets in the entire data set. For example, 80% of targets covered in the top 20% of data based on the model.

The lift measures how much better one can expect to do with the predictive model vs without a model. It is the ratio of gain % to the random expectation % at a given decile level.

 

Assets Included

  1. R notebooks
  2. Sample Files
  3. Cognos Dashboards

Prerequisites of Cloud Pak for Data Services 

  1. R Studio
  2. Watson Machine Learning
  3. Watson Studio
  4. Cognos Analytics

Enablement Material

- Please refer here - Box Folder

Import the Accelerator

- Please watch the video available in the box folder for the use case. For any questions or clarifications, please contact ddp.helpdesk@wipro.com

Release Notes

-This has been tested with CPD Version 2.1

About the Developer

Wipro Limited (NYSE: WIT, BSE: 507685, NSE: WIPRO) is a leading global information technology, consulting and business process services company. We harness the power of cognitive computing, hyper-automation, robotics, cloud, analytics and emerging technologies to help our clients adapt to the digital world and make them successful. A company recognized globally for its comprehensive portfolio of services, strong commitment to sustainability and good corporate citizenship, we have over 160,000 dedicated employees serving clients across six continents. Together, we discover ideas and connect the dots to build a better and a bold new future.

 


#Featured-area-2
#datascience
#banking
#IndustryAccelerator

Statistics

0 Favorited
68 Views
0 Files
0 Shares
0 Downloads