AI glossary Strategy & Business
What is predictive analytics?
← All glossary termsUsing historical data, statistics and machine learning to estimate what is likely to happen next: demand, churn, failures, fraud or sales.
Explained
Predictive analytics uses historical data together with statistical and machine learning models to estimate future outcomes or unknown values. Typical questions are how much demand there will be next month, which customers are likely to leave, which machine is likely to fail or which transaction looks fraudulent. The output is usually a probability or a forecast, which is then used to act earlier: adjust stock, contact a customer or schedule maintenance. It sits between descriptive analytics, which explains what happened, and prescriptive analytics, which recommends what to do.
How it works
- Define the outcome to predict and how the prediction will be used.
- Gather historical data where that outcome is known, and build the input variables.
- Train a model, such as regression, gradient-boosted trees or a neural network, and validate it on data it has not seen.
- Deploy it into the process where decisions are made, not only into a report.
- Monitor accuracy over time and retrain when the data or the business changes.
Common uses
- Demand and sales forecasting.
- Customer churn and propensity to buy.
- Predictive maintenance of equipment.
- Fraud and credit risk scoring.
- Staffing and capacity planning in contact centres or hospitals.
- Claims severity and reserve estimation in insurance.
What makes it work
The model is rarely the hard part. What decides the result is whether there is enough reliable historical data, whether the prediction arrives in time to change a decision and whether someone is accountable for acting on it.
Models also degrade as conditions change, so monitoring for drift is part of the work. And when predictions affect people, such as credit or insurance decisions, fairness, explainability and regulation must be considered from the start.
Frequently asked questions
What is predictive analytics in simple terms?
It is using what has happened in the past to estimate what is likely to happen next. A model learns patterns from historical data and applies them to current cases to produce a forecast or a probability.
What is the difference between predictive and prescriptive analytics?
Predictive analytics estimates what is likely to happen. Prescriptive analytics goes a step further and recommends the best action, often by combining predictions with optimisation and business rules.
Is predictive analytics the same as machine learning?
Not exactly. Machine learning is one of the main techniques used in predictive analytics, alongside classical statistics. Predictive analytics is the business discipline; machine learning is part of its toolkit.
What data do you need for predictive analytics?
Historical records where the outcome you want to predict is known, plus the variables that might explain it. The data must be reasonably complete, consistent over time and available when the prediction is needed.
How accurate is predictive analytics?
It depends on the problem, the data and how stable the conditions are. A model should always be validated on data it has not seen and compared with a simple baseline, such as last year's figures, before it is trusted.
How to put it to work
Where it shows up at Thinkia
Further reading
Related terms
Sources
Updated: 2026-10-21