Strategic Business Leader · Big data and data analytics

Types of Data Analytics and Techniques for SBL

Updated 11 October 2026 · Fact-checked

Data analytics is the use of data to gain insight for decisions. Four types build on each other: descriptive (what happened), diagnostic (why), predictive (what will happen) and prescriptive (what should we do). In SBL, name the type, apply it to the scenario, and weigh benefits against risks.

Understand Data Analytics Types and Techniques

Data analytics means examining data to find patterns and support decisions. Big data gives the raw material: large, fast and varied data. Analytics turns that material into insight a manager can act on.

There are four types, and each answers a harder question than the last:

Descriptive analytics: what happened? It summarises past data, for example monthly sales by region or website visits. Dashboards and reports are typical outputs.
Diagnostic analytics: why did it happen? It digs into causes, for example why sales fell in one region. It uses drill-down, comparison and correlation.
Predictive analytics: what is likely to happen? It uses historical data and statistical models to forecast, for example which customers may leave or what demand will be next quarter.
Prescriptive analytics: what should we do? It recommends actions, often by testing options against constraints, for example the best price, route or stock level.

Value and difficulty rise as you move down the list. Descriptive is easy and common but only looks backwards. Prescriptive is powerful but needs good data, strong models and trust in the output.

Two techniques are often mentioned. Data mining searches large data sets for hidden patterns and relationships, such as products that customers buy together. Machine learning is where software learns from data and improves its predictions without being given fixed rules for each case. It often powers predictive and prescriptive work, such as fraud detection or recommendation engines.

In SBL, you are not asked to build models. You are asked to advise a board. Explain what type of analytics fits the problem, what data it needs, what benefit it gives, and what risks to manage: poor data quality, privacy and ethics, bias, cost, and over-reliance on a model.

Key rules to remember

Descriptive analytics: What happened? (past data, summaries); Reports, dashboards, averages and trends. Hindsight only.
Diagnostic analytics: Why did it happen? (causes, drill-down); Finds reasons behind results. Correlation does not prove cause.
Predictive analytics: What is likely to happen? (models, forecasts); Uses past patterns to estimate future outcomes. Results are probabilities, not certainties.
Prescriptive analytics: What should we do? (recommended actions); Suggests the best option given goals and constraints. Needs human judgement and oversight.
Data mining vs machine learning: Data mining = finding patterns; machine learning = learning from data to improve predictions; Data mining discovers patterns. Machine learning builds models that adapt as more data arrives.

How to solve Data Analytics Types and Techniques questions

Use this method for any SBL requirement on analytics, whether it asks you to explain, recommend or evaluate.

1Read the requirement and note the verb: explain, recommend, evaluate or advise. It sets the depth and the balance of your answer.
2Identify the business problem in the scenario: for example falling margins, customer losses or poor forecasting.
3Match the problem to the analytics type: past, cause, future or action. Name the type explicitly.
4Say what data is needed and where it comes from: internal systems, customers, sensors, social media or external sources.
5Explain the benefit in terms of the scenario: better decisions, cost savings, personalised offers, earlier warnings.
6Cover the risks and limits: data quality, privacy law, ethics, bias, cost, skills and over-reliance on models.
7Give a clear recommendation with a sensible first step, such as a pilot, and who should own it.
8Check you have used scenario facts and a professional tone, since professional skills marks depend on it.

Quickest way: Question-Type-Benefit-Risk

When to use it: Use when time is short and you need a structured paragraph quickly.

Question: write the business question in the scenario (what, why, what next, what to do).
Type: name the matching analytics type in one line.
Benefit: give one or two scenario-specific gains.
Risk: give one or two risks and a control for each.
Close with one clear recommendation.

Common mistakes in Data Analytics Types and Techniques

Listing the four types as a definition with no link to the scenario.
Students memorise the theory and stop there.
Fix: For each type you mention, tie it to a problem or data source in the case and state the benefit.
Confusing diagnostic with descriptive analytics.
Both look at past data.
Fix: Descriptive says what happened. Diagnostic explains why. If the question asks for causes, it is diagnostic.
Treating predictive output as certain.
Models look scientific and precise.
Fix: Say forecasts are probabilities that depend on data quality and stable conditions, and recommend human review.
Ignoring ethics and data protection.
Students focus on the technology benefits.
Fix: Always add a risk point on privacy, consent, bias and transparency, especially when personal data is used.
Saying data mining and machine learning are the same thing.
Both are used with large data sets.
Fix: Data mining finds patterns in data. Machine learning builds models that learn and improve. They often work together.
Recommending analytics with no practical steps.
Students stop at advantages.
Fix: Add implementation points: data quality checks, skills, a pilot, governance and measuring results.

Worked examples

Example 1

A retail chain has seen online sales fall for three months. The board asks you to explain how the four types of analytics could help it understand and respond. (10 marks style)

Show the solution

Descriptive: summarise sales by product, region, channel and month to show where the fall is happening, for example one product category or one region.
Diagnostic: drill into that area to find why. Compare with pricing, website visits, conversion rates, delivery times and competitor moves. Customer reviews may show a cause.
Predictive: use past customer behaviour to forecast which customers are likely to stop buying and what demand may be if nothing changes.
Prescriptive: test options such as targeted discounts, changed stock levels or delivery improvements, and recommend the one that best meets margin and sales goals.
Risks: poor or incomplete data, privacy rules on customer data, and over-reliance on models. Recommend a pilot in one region with management review of results.

Answer: Descriptive shows where sales fell, diagnostic explains why, predictive estimates what will happen next, and prescriptive recommends what to do. The retailer should start with a pilot, check data quality, protect customer data and keep managers in the decision.

Example 2

A bank wants to use machine learning to spot fraudulent card transactions. Advise the board on the type of analytics involved, the benefits and the key risks. (8 marks style)

Show the solution

Type: this is mainly predictive analytics. The model learns from past transactions marked as fraud or genuine and estimates the likelihood that a new one is fraud.
Technique: data mining can find unusual patterns in transaction history. Machine learning then improves the model as new cases arrive.
Benefit: faster detection, lower fraud losses, fewer manual checks and better customer protection.
Possible prescriptive step: the system can recommend or trigger an action, such as blocking the card or sending a verification request.
Risk: false positives block genuine customers and harm trust. Poor or biased training data gives poor results. Customer data brings privacy and regulatory duties. Fraudsters change tactics, so the model needs regular review.
Recommendation: use the model with human review of high-value alerts, monitor accuracy, and document how decisions are made.

Answer: The bank is using predictive analytics, supported by data mining and machine learning, with prescriptive actions on alerts. Benefits are lower losses and faster detection. Risks are false alarms, biased or poor data, privacy duties and changing fraud methods, so it should keep human oversight and review the model regularly.

Exam tips

Link every type of analytics to a problem in the case. Generic definitions earn few marks.
Use the verb in the requirement. Evaluate needs benefits and limits. Recommend needs a clear decision.
Always include a risk and ethics point, such as privacy, bias or poor data quality.
Keep answers in short, labelled paragraphs. This helps clarity and professional skills marks.
Remember analytics supports judgement. Say that managers must still challenge the output.

Practice questions from Big data and data analytics

Data Analytics Types and 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 Types and Techniques: frequently asked questions

What are the four types of data analytics?

They are descriptive, diagnostic, predictive and prescriptive. They ask what happened, why it happened, what will happen and what to do. Each type adds more insight and needs more skill and data.

What is the difference between predictive and prescriptive analytics?

Predictive analytics estimates what is likely to happen. Prescriptive analytics goes further and recommends the best action given goals and constraints. Prescriptive output still needs human judgement.

How is machine learning different from data mining?

Data mining searches data for patterns and relationships. Machine learning builds models that learn from data and improve their predictions over time. In practice they are often used together.

How should I use data analytics in an SBL answer?

Identify the business problem, name the matching analytics type, and explain the data needed and the benefit in the scenario. Then cover risks such as data quality and privacy, and finish with a practical recommendation.