ML-based Consumer Segmentation Strategy Leads to 15% Improvement in Marketing Spend Optimization
Automated the creation of customer profiles, created segmentation strategy and recommended the best profiles to be targeted across marketing channels using ML
The customer manufactures and sells consumer lawn, garden, and pest control products and wanted to maximize customer loyalty and acquire new customers by leveraging segmentation strategy by segmenting easily acquirable customers and those with high lifetime value. The existing process of collecting and analyzing customer data was manual and marketing campaigns were designed based on intuition leading to low conversions and high marketing spending. The customer wanted to automate the process of collecting consumer data and build models to predict conversion rates and cluster them into segments to plan effective marketing campaigns.
Using historical data, Sigmoid built two machine learning prediction models — the “acquisition model”, which predicted the probability of acquisition, and ‘LTV model’ which predicted the lifetime value of a customer. Once the two models were created, we took the trained model and applied SHAP analysis to find out the top features and key drivers impacting the models, such as age, demographics, etc. We used these top features to create customer profiles. The output of the analysis was displayed on dashboards for efficient analysis by the marketing team. The solution also recommended top three profiles across each of the websites for efficient targeting.
Using the new customer profiles, the business was able to significantly reduce the number of redundant campaigns and create tailored marketing communications for easily acquirable and high lifetime value customers across various marketing channels.
Improvement in marketing spend optimization
Increase in Life Time Value of customers for different brands
More acquisition of customers for different brands across marketing channels