Author: Dr Angel Salazar
Short bio:
Folliowing a career as a lecturer at business schools, that included working in business intelligence and predictive marketing analytics, I am working on the 'Full-Stack Customer eXperience Analytics and AI' framework. I have overseen projects including e-commerce, operations and logistics, and marketing strategy, comms and analytics. I have developed new training on customer analytics, using tools such as Excel, PowerBI, Python and Pandas, performing a range of analyses. These include multi-variate regression and econometric methods:
- Customer segmentation using personas, K-means clustering, product recommender systems, basket analysis, likelihood of purchase and churn, lifetime value, and marketing mix analysis.
- Multivariate regression modelling, demand analysis and choice modelling, and econometric methods.
This repository will provide a summary of core concepts and techniques for analysing and automating customer experience and customer value within a commercial or business context.
The overriding question is how can data scientists ensure that they are doing all things necessary to support the business and serve their customer better.
Without dweling too much on marketing theory, data scientists need a minimum subject knowledge to be able to answer the questions below in order to develop and deploy useful machine learning models:
What are the key customer segments? Why are they important for maximising sales?
What is their customer value lifecycle? Why is it important?
What is a customer experience journey? How does it relate to the experience journey?
What are the key challenges and important decisions that managers need to take into consideration at each stage of the customer journey and value lifecycle?
- Which customers would respond to our adverts and promotions?
- Which customers are not satisfied and likely to leave?
- What products are they likely to purchase, are they likely to repeat their purchases?
What data do businesses need to gather, transform and inject into their machine learning models?
How should this data be handled, cleaned and transformed so it is ready for use?
Which are the most suitable machine learning methods to analyse all these questions above?
What are the steps that data scientists need to follow to ensure accurate and repeatable outcomes?
How should machine learning models be maintained and extended?
What parts of the data analytics lifecycle can be automated to ensure cost-effective operations?
The framework intends to provide a clear structure and approach to customer-centric marketing analytics using data science and machine learning techniques.
Folders with marketing analytics coding applications to follow shortly ...