Machine Learning for Beginners – ChatGPT as a Data Analyst for Your Marketing

Introduction to Machine Learning and a practical demo: How ChatGPT acts as a free data analyst, carrying out regression analysis and customer segmentation — without any programming knowledge.

What you will learn

Many companies have a lot of data, but do very little with it. ChatGPT solves this: you have a free data analyst at your disposal.

Analysis 1: Revenue forecast via regression analysis

Dataset: TV, radio and newspaper spend versus resulting revenue.

Approach: upload the CSV to ChatGPT, let it explore the data, calculate a linear regression, visualise the result, and have an interactive budget simulator interface programmed.

Result: 90% explained variance. TV and radio have a much greater influence than print. The generated code can be passed directly to developers.

Analysis 2: Customer segmentation via K-means clustering

Dataset: over 2,000 customers with demographic data, purchases, revenue, promotion responses and purchase channel.

Approach: clean the data (delete unrealistic values), run K-means with the elbow method (4 segments recommended), generate cluster descriptions, and produce marketing recommendations for each segment.

Example segments: solid buyer (older, buys wine/meat), affluent female customers (premium products), low-price segment (gamification recommended).

Key prompting tips

Opening prompt: You are a data analyst. Briefly explain the dataset. Proceed step by step, and wait for my go-ahead.
Guide it step by step. Do not let it run everything at once.
Explain cryptic field names in the prompt
Ask for code generation directly: interactive interface, automatic customer assignment