Machine Learning.

What is machine learning?

Machine learning is a branch of artificial intelligence in which software learns patterns from data and uses them to make predictions or decisions, instead of following rules a person wrote by hand.

Show a system thousands of past examples, such as which leads became customers, and it learns the signals that tend to come before a sale. Generative AI is one recent offshoot; much machine learning quietly scores, sorts and forecasts.

Why it matters

You already rely on machine learning. Spam filters, fraud checks on card payments, Google Ads Smart Bidding and Performance Max, and product recommendations all run on it. The commercial point is that these systems learn from the data you feed them. Send an ad platform weak signals, such as page views counted as conversions, and it will optimise hard towards the wrong thing. The quality of your tracking and records sets the ceiling on how well the machine performs.

The main types

• Supervised learning The system learns from labelled examples, such as emails marked spam or not spam, then labels new ones.
• Unsupervised learning It finds groupings in unlabelled data, such as clusters of similar customers.
• Reinforcement learning It learns by trial and feedback, adjusting its actions to earn better results over time.
• Deep learning Large neural networks that handle images, speech and language, and underpin today's generative AI.

What to do about it

Check whether your CRM records outcomes consistently: won or lost, the value and the reason. If it does not, start now, because every month of clean data makes future prediction possible. For paid advertising, make sure the conversions you track are real enquiries or sales, and where possible send qualified leads and closed deals back to the ad platform so its bidding learns what a good customer looks like.