Performance Management · Big data and data analytics

Data Analytics Techniques: Descriptive, Diagnostic, Predictive, Prescriptive

Updated 11 October 2026 · Fact-checked

Data analytics is the process of examining data to find patterns and support decisions. It has four types. Descriptive shows what happened. Diagnostic shows why. Predictive shows what is likely to happen. Prescriptive recommends what to do. To answer exam questions, match each type to its question and its example.

Understand Data Analytics Techniques

Data analytics means examining data to draw conclusions and help people decide. It is not the same as big data. Big data describes the data itself: very large, fast and varied sets of data (the Five Vs). Data analytics is what you do with data, big or small, to get useful insight. You can run analytics on a small spreadsheet.

The four types form a ladder. Each step answers a harder question and usually adds more value, but also needs more skill, data and technology.

Descriptive analytics answers: what happened? It summarises past data using totals, averages, charts and dashboards. Example: monthly sales by region.
Diagnostic analytics answers: why did it happen? It digs into the data to find causes, using drill-down, comparing groups and looking for correlations. Example: finding that sales fell because one product was out of stock in the north.
Predictive analytics answers: what is likely to happen? It uses past patterns, statistics and forecasting models to estimate future outcomes. Example: forecasting next quarter's demand or the chance a customer will leave.
Prescriptive analytics answers: what should we do? It recommends actions, often by testing many options with models, simulation or optimisation. Example: setting the best price or the best delivery routes.

Management accountants use all four. Variance reports are descriptive. Investigating a variance is diagnostic. Forecasts and budgets built on trends are predictive. Choosing the production mix or a price that maximises profit is prescriptive.

Analytics is only as good as the data behind it. Poor quality, biased or incomplete data gives misleading results. A correlation does not prove one thing causes another, so a good analyst checks findings with judgement.

Key rules to remember

Descriptive analytics: Question: What happened?; Summarises past data. Typical outputs: reports, dashboards, averages, totals.
Diagnostic analytics: Question: Why did it happen?; Finds causes. Typical methods: drill-down, comparison, correlation, variance investigation.
Predictive analytics: Question: What is likely to happen?; Estimates future outcomes using past data and models. Gives likelihoods, not certainty.
Prescriptive analytics: Question: What should we do?; Recommends the best action, often using optimisation or simulation. Needs the most data and skill.
Data analytics vs big data: Big data = the data; data analytics = the analysis of data; Analytics can be applied to data of any size.

How to solve Data Analytics Techniques questions

Use this method for any question that asks you to identify, explain or apply the types of data analytics.

1Read the scenario and underline what the organisation is actually doing or wants to know.
2Turn it into a question: what happened, why, what next, or what should we do?
3Match the question to the type: descriptive, diagnostic, predictive or prescriptive.
4Check for a trap: reporting past results is descriptive even if the data is huge; explaining a cause is diagnostic, not predictive.
5For written answers, define the type, then link it to the scenario with a specific example.
6Add the benefit to the business, such as faster decisions, better forecasts or lower cost.
7Add one limitation if the question asks for evaluation, such as poor data quality, cost, skills shortage or false correlations.

Quickest way: Four-question shortcut

When to use it: Use it for Section A and Section B objective test questions where you must name the type of analytics.

Look at the verb in the scenario: report or summarise means descriptive.
Why, cause or investigate means diagnostic.
Forecast, likely or probability means predictive.
Recommend, optimise or best option means prescriptive.
If two seem to fit, choose the highest step the scenario fully achieves.

Common mistakes in Data Analytics Techniques

Treating data analytics and big data as the same thing.
They appear together in the syllabus and news stories.
Fix: Say big data is the data (volume, velocity, variety and so on) and analytics is the analysis of it. Analytics works on small data too.
Calling a forecast prescriptive.
Both look to the future.
Fix: Predictive says what will probably happen. Prescriptive goes further and recommends an action.
Labelling a variance report as diagnostic.
Variances feel like analysis.
Fix: Reporting the variance is descriptive. Finding out why it arose is diagnostic.
Giving a definition with no link to the scenario.
Students recall theory but skip application.
Fix: Use the company's own data, product or customers in every point.
Assuming analytics gives certain answers.
Computer output looks precise.
Fix: Note that models rely on data quality and assumptions, and a correlation does not prove causation.
Listing the types but not explaining the benefit.
Students stop once the types are named.
Fix: State what decision each type supports and how it improves performance.

Worked examples

Example 1

A retailer reviews last quarter's sales and finds that sales in its northern stores fell by 12%. It then examines stock records and finds that its best-selling product was out of stock in those stores for six weeks. It now builds a model to estimate next quarter's demand by store. Identify the type of analytics at each stage.

Show the solution

Stage 1: reviewing the fall in sales answers 'what happened?'. This is descriptive analytics.
Stage 2: examining stock records to find the cause answers 'why?'. This is diagnostic analytics.
Stage 3: building a model to estimate next quarter's demand answers 'what is likely to happen?'. This is predictive analytics.
No stage yet recommends an action, so prescriptive analytics has not been used.

Answer: Stage 1 is descriptive, stage 2 is diagnostic and stage 3 is predictive. Prescriptive analytics would be a further step, such as a model recommending the stock level for each store.

Example 2

A delivery company has a large set of data on routes, traffic and fuel use. It wants software to recommend the daily routes that minimise cost. Explain which type of analytics this is and give two benefits and one limitation.

Show the solution

The company wants a recommended action, the lowest-cost routes. This answers 'what should we do?'.
That is prescriptive analytics. It typically tests many options using optimisation or simulation.
Benefit 1: it can find lower-cost routes than manual planning, reducing fuel and driver costs.
Benefit 2: it can update recommendations as traffic changes, giving faster and more consistent decisions.
Limitation: the output depends on accurate, complete data and correct assumptions. Poor data gives poor recommendations, and the system may be costly and need specialist skills.

Answer: This is prescriptive analytics. Benefits are lower costs and faster, consistent decisions. A limitation is reliance on data quality and the cost and skills needed to run it.

Exam tips

Learn the four questions (what, why, likely, should) and use them to classify any scenario.
In written answers always tie each type to the scenario, not to a generic example.
Be ready to explain the difference between data analytics and big data in one or two sentences.
For a management accounting angle, link types to familiar tools: variance reports, forecasts, budgets and decision models.
If asked to evaluate, give both benefits and limitations, including data quality and skills.

Practice questions from Big data and data analytics

Data Analytics Techniques in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Data Analytics Techniques: frequently asked questions

What are the four types of data analytics?

They are descriptive, diagnostic, predictive and prescriptive. They answer what happened, why it happened, what is likely to happen and what should be done. Each step usually adds more value but needs more data and skill.

What is the difference between data analytics and big data?

Big data refers to the data itself, which is very large, fast-moving and varied. Data analytics is the process of analysing data to find insight. You can use analytics on small data sets as well.

How do management accountants use data analytics?

They use it to report performance, investigate variances, forecast demand and costs, and choose between options such as pricing or production plans. It helps them give faster and better-supported advice to managers.

Is predictive analytics the same as prescriptive analytics?

No. Predictive analytics estimates what is likely to happen. Prescriptive analytics recommends what action to take, often by comparing many possible options.