Data science guide
A non-technical guide to advanced analytics, algorithms, and ways of working.
What this guide covers
Advanced analytics goes beyond reporting and dashboards. It uses algorithms and models to predict, segment, recommend, or explain what is happening in your data. This guide explains the main approaches in plain language, the work involved in using them, and how agentic analytics differs from traditional human-led analytics.
The guide is in five parts:
- What is advanced analytics? — Descriptive, diagnostic, and advanced analytics; what algorithms are and why they matter.
- Kinds of algorithms (in plain language) — Regression, classification, clustering, time series, NLP, and optimisation.
- The main tasks in building advanced analytics — From defining the problem to deployment and maintenance.
- Agentic vs traditional analytics — Who does the work, speed, scale, and where people focus.
- Analytics patterns — How different types of questions map to analytics approaches, with worked examples.
Where to go from here
- To plan which advanced analytics to build and in what order, see TrueState 360.
- To roll out agentic analytics in your organisation, see Deploying world-class analytics (change management).