Advanced Performance Management · Management reports

Big Data, Analytics and Information Quality in APM Management Reports

Updated 11 October 2026 · Fact-checked

Big data is very large, fast, varied data that normal tools cannot handle. Analytics turns it into information for management reports. In APM you explain how it improves reporting, then judge the information's quality (accuracy, completeness, timeliness, relevance, reliability) and state limitations, applying all of this to the scenario.

Understand Big Data, Analytics and Information Quality

Data is raw facts and figures with no context, such as a list of transaction amounts. Information is data that has been processed, organised and put in context so a manager can use it to decide or act. A management report is a vehicle for information. Poor data gives poor information, however good the report looks.

Big data is usually described by the 'V's: volume (very large amounts), velocity (generated and processed at speed), variety (structured data like sales ledgers and unstructured data like social media posts, emails, images and sensor feeds). Some lists add veracity (how trustworthy the data is) and value. Big data can come from inside the organisation or from outside, such as web traffic, customer reviews and market feeds.

Data analytics is the process of examining data to find patterns and support decisions. Four types are commonly used: descriptive (what happened), diagnostic (why it happened), predictive (what is likely to happen) and prescriptive (what should we do). For management reports, this means more timely dashboards, richer KPIs, forecasts, and the ability to drill down into causes rather than just report totals.

But more data is not better information. Reports need information that is relevant, accurate, complete, timely, understandable, and obtained at a cost below its benefit. Reliability depends on the source, how the data was captured and cleaned, and how it was analysed. Common limitations are errors and bias in the data, correlation mistaken for causation, information overload, cost of systems and skills, cyber and privacy risks, and over-reliance on what is easy to measure.

In the exam, the marks come from application. Link the technology to the business in the scenario: what data it has, what decisions managers need to make, and what could go wrong with the information they would receive.

Key rules to remember

Data to information: Data + processing + context = Information; Use this to answer 'difference between data and information' questions. Add that information must be useful to a decision.
Big data characteristics: Volume, Velocity, Variety (plus Veracity and Value in extended lists); Name the ones you use and apply each to the scenario, not just list them.
Types of analytics: Descriptive → Diagnostic → Predictive → Prescriptive; Moves from what happened to what should be done. Value and difficulty usually rise along the sequence.
Qualities of good information: Accurate, Complete, Timely, Relevant, Understandable, Cost-effective (and reliable source); Many memory aids exist (such as ACCURATE). Any sensible, well-explained list earns credit.
Cost-benefit test: Value of better decisions > cost of obtaining the information; Applies to any new data source or system. Benefits are often hard to quantify.

How to solve Big Data, Analytics and Information Quality questions

Use this method for any question on big data, analytics or information quality in management reports.

1Read the requirement and note the verb: explain, assess, evaluate, advise or recommend. Note who the audience is (board, manager, regulator).
2Pick out scenario facts: the industry, the data the entity holds or could collect, current reporting weaknesses and the decisions managers face.
3Define key terms briefly (data, information, big data, a type of analytics) and move quickly to application.
4Explain the benefits to management reporting, tied to the scenario: timeliness, deeper insight, better forecasts, more relevant KPIs.
5Assess quality and reliability: check the source, accuracy, completeness, timeliness, bias, and whether the measure is relevant to the decision.
6Cover limitations and risks: cost, skills, overload, data protection, cyber risk, false patterns and reliance on past data.
7Give a reasoned recommendation or conclusion, such as a phased approach, data governance or pilot, and show professional scepticism.
8Check that each point is linked to the scenario and the requirement, not a generic list.

Quickest way: Benefit, quality, limitation (BQL) plan

When to use it: When you have about 10 minutes for a short written requirement and need a structure fast.

Write three headings on your plan: Benefit, Quality, Limitation.
Under Benefit, note two scenario-specific uses of the data or analytics.
Under Quality, note two tests of reliability that matter most in the scenario, such as source and completeness.
Under Limitation, note two risks, such as cost and privacy or cyber risk.
Add one recommendation, then write, giving each point a reason ('because...') linked to the scenario.

Common mistakes in Big Data, Analytics and Information Quality

Writing a generic list of the Vs of big data with no link to the business.
Students memorise the list and assume definitions earn the marks.
Fix: Define in one line, then apply each V to the scenario's data, such as velocity of live sales feeds in a retailer.
Treating data and information as the same thing.
Both words are used loosely at work.
Fix: State that information is processed, contextualised data useful for a decision, and show the processing step in your answer.
Saying more data always gives better reports.
Technology is assumed to be an improvement automatically.
Fix: Discuss overload, noise, poor quality and cost. Say that value depends on relevance and reliability, not volume.
Ignoring reliability and bias in the data source.
Students focus on the technology's benefits and skip the evaluation.
Fix: Ask who collected the data, how, when, and whether it covers all customers or only a biased group, such as online reviewers.
Confusing correlation with causation in analytics.
Patterns in large datasets look convincing.
Fix: Say a pattern needs testing and business logic before it drives decisions or targets.
Leaving out data protection, security and ethics.
These feel like IT topics, not performance management.
Fix: Include a short point on privacy law compliance, cyber risk, and the reputational damage from misusing personal data.

Worked examples

Example 1

A retail chain with 200 stores produces monthly management reports from its accounting system. The board wants to use customer loyalty-card data, website activity and social media comments to improve reporting. Explain two benefits and two limitations of this approach for management reporting. (8 marks, technical)

Show the solution

Define briefly: this is big data, as it is large in volume, arrives fast (velocity) and mixes structured sales data with unstructured comments (variety).
Benefit 1: timeliness and detail. Loyalty-card and web data can be reported weekly or daily by store and product, rather than monthly totals. Managers can react to stock-outs or falling sales sooner.
Benefit 2: richer insight. Analytics can show why sales differ between stores (diagnostic) and forecast demand (predictive). Reports could then include customer-based KPIs, such as repeat purchase rates, not only financial results.
Limitation 1: quality and reliability. Loyalty-card holders may not represent all customers and social media comments may be biased or unverified. Reports built on them could mislead.
Limitation 2: cost, risk and overload. Systems, skills and storage are costly, and personal data brings privacy law and cyber risk. Managers may also face too many measures and ignore the key ones.
Conclude: the chain should start with a pilot in selected stores, set data quality checks and governance, and keep reports focused on a few relevant KPIs.

Answer: Benefits: faster, more detailed reporting and deeper diagnostic and predictive insight. Limitations: unrepresentative or unreliable data, and cost, privacy, cyber and overload risks. Recommend a governed pilot with a small set of relevant KPIs.

Example 2

A manufacturer's operations report shows machine downtime, which the board uses to set bonuses for plant managers. The data is typed in by supervisors at the end of each week. Assess the reliability of this information and recommend improvements. (8 marks, including professional skills)

Show the solution

Identify the issue: the report is used for rewards, so managers have an incentive to bias the data. This matters for reliability.
Accuracy: manual entry at week end relies on memory and is open to error or deliberate understatement of downtime.
Timeliness: weekly data arrives too late to fix problems when they occur.
Completeness and relevance: downtime alone ignores causes, planned maintenance and output quality. It may not link to the plant's objectives, so the report could encourage managers to keep machines running at the cost of quality.
Show scepticism: ask who prepares, checks and benefits from the data. There is no independent verification.
Recommend: install sensors that record downtime automatically (big data from the Internet of Things), apply analytics to find causes and predict failures, add independent checks or internal audit sampling, and report downtime with quality and output measures. Keep bonuses from relying on one measure.

Answer: The information is of low reliability: manual, late, open to bias and narrow in scope. Automatic sensor capture, analytics on causes, independent review and a balanced set of measures would improve reliability and make the report more useful for decisions.

Exam tips

Always tie each point to the scenario's industry, data and decisions. Generic definitions earn few marks.
Balance your answer: when asked to evaluate, give benefits, limitations and a conclusion, not just one side.
Use the professional skills marks: structure with clear headings, write for the stated reader, and show scepticism about data sources.
Do not spend long on definitions. Give one line, then analyse. Examiners reward analysis and judgement.
Mention data protection, cyber risk and ethical use in at least one point where personal data appears in the scenario.

Practice questions from Management reports

Big Data, Analytics and Information Quality in other exams

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

Big Data, Analytics and Information Quality: frequently asked questions

What is the difference between data and information in management reporting?

Data is raw, unprocessed facts. Information is data that has been processed, organised and put in context so that it helps a manager make a decision. A report should contain information, not just data.

What makes information reliable in management reports?

Reliable information comes from a trustworthy source, is captured and processed accurately, is complete and up to date, and is free from bias. Independent checks and clear data governance improve reliability. Information can be accurate but still not relevant to the decision.

How does data analytics improve performance reporting?

It allows faster, more detailed reports and helps explain results, forecast and recommend actions. Descriptive analytics shows what happened, diagnostic shows why, predictive shows likely outcomes and prescriptive suggests actions. Its value depends on the quality of the data.

What are the main limitations of big data for management reports?

The main limitations are poor or biased data, cost and skills, information overload, false patterns from correlation, and privacy and cyber risks. Past data may also not predict a changed future. Mention the ones that fit the scenario.