vaayaa Inclusive Education
Technical risk assessment

Business Intelligence Management Risk Assessment: Data Quality, Privacy and Analytical Controls

A worked example of how a business intelligence team identifies, scores and controls risks from licensed data, sampling bias, privacy gaps, pipeline failure and misleading dashboards, with owners and review triggers.

Updated November 2026 8 hazards assessed Review every 6 months Highest risk: Critical to Medium
See the Business Intelligence Management course

Executive summary

This assessment shows how a business intelligence team can find, rate and control the risks that determine whether its dashboards and models can be trusted. Eight risks were chosen because they are plausible in everyday analytics work, from licensed data and personal data to pipelines, reproducibility and the ethical use of findings. Privacy and consent carry the highest inherent score because a breach affects people directly and triggers enforcement. Sampling bias, licence misuse and misleading dashboards follow closely. Each risk is scored before controls and again after existing and added controls. Most fall to Medium or Low once controls are in place, while privacy stays at High because its impact remains serious. The first actions are a privacy review of extracts, a licence tagging exercise, a row count threshold for loads and a versioning rule for published outputs. The organization in this example is a teaching device, so every score illustrates method rather than describing a real team.

Focus area
Business intelligence management and analytics governance
Setting
Analytics function of a mid-sized organization that serves business leaders from a shared data platform and dashboards
Method
Five by five likelihood and impact risk register

Scope

This assessment covers a business intelligence function that collects data from internal systems and licensed third-party sources, transforms it, and publishes dashboards, reports and analytical models to managers. It covers data quality, sampling, privacy, access, retention, reproducibility, pipeline reliability and the ethical use of findings. Jurisdiction, data volumes, tools and contract terms are not specified and are treated as assumptions for teaching.

Context

The team has a head of analytics, a data engineering lead, several analysts, a data steward for each major source and a privacy officer who advises on personal data. Data arrives through scheduled loads into a central warehouse, where it is joined, cleaned and presented in dashboards and models. Some datasets are licensed from outside suppliers under contract. Each risk is scored before controls to show its natural level, then again after the existing controls and the additional controls that should be added.

Objectives, assets and stakeholders

The objective is to deliver timely, accurate and lawful insight that managers can rely on, while protecting personal data, honoring supplier licence terms and keeping analysis reproducible. The assets in scope are the warehouse and its data, licensed datasets, personal records and consent information, dashboards and models, code and documentation, and the reputation of the analytics function. Stakeholders include analysts, data stewards, managers who act on findings, customers and employees whose data is used, suppliers who license datasets, the privacy officer and the governance committee. Every risk has one owner role, and the head of analytics keeps the register current and reports to the governance committee each quarter.

Risk matrix: before and after controls

Each cell shows hazard numbers placed by likelihood and impact. Compare the two grids to see how controls move each hazard.

Before controls

Inherent risk

Risk before controls: rows are likelihood from almost certain to rare, columns are impact from insignificant to catastrophic. Cells list hazard numbers.
1
Insignificant		 2
Minor		 3
Moderate		 4
Major		 5
Catastrophic
5 Almost certain
4 Likely Likelihood 4, impact 3, score 12, high: H05 Likelihood 4, impact 4, score 16, high: H01 H02 Likelihood 4, impact 5, score 20, critical: H03
3 Possible Likelihood 3, impact 3, score 9, medium: H07 Likelihood 3, impact 4, score 12, high: H04 H06
2 Unlikely Likelihood 2, impact 4, score 8, medium: H08
1 Rare

After controls

Residual risk

Risk after controls: rows are likelihood from almost certain to rare, columns are impact from insignificant to catastrophic. Cells list hazard numbers.
1
Insignificant		 2
Minor		 3
Moderate		 4
Major		 5
Catastrophic
5 Almost certain
4 Likely
3 Possible
2 Unlikely Likelihood 2, impact 2, score 4, low: H06 H07 Likelihood 2, impact 3, score 6, medium: H01 H02 H04 H05 Likelihood 2, impact 4, score 8, medium: H03
1 Rare Likelihood 1, impact 3, score 3, low: H08
  • Low 1 to 4
  • Medium 5 to 9
  • High 10 to 16
  • Critical 17 to 25

Hazard register

A summary first, then the detail for each hazard.

Summary of hazards with scores before and after controls
No.HazardBeforeAfterOwner
H01 Breach of third-party data licensing terms 16 High 6 Medium Procurement and contracts lead
H02 Sampling bias and poor source data quality 16 High 6 Medium Data quality lead
H03 Unlawful use of personal data and consent gaps 20 Critical 8 Medium Privacy officer
H04 Excess access rights and data kept beyond need 12 High 6 Medium Data steward
H05 Misleading dashboards and biased analytical models 12 High 6 Medium Analytics manager
H06 Pipeline failure and stale reporting 12 High 4 Low Data engineering lead
H07 Irreproducible analysis and weak methodology 9 Medium 4 Low Analytics manager
H08 Misuse of insights and missing ethical review 8 Medium 3 Low Governance committee chair
H01

Breach of third-party data licensing terms

Legal and compliance Before 16 After 6

Licensed market or benchmark datasets are used beyond the permitted purpose, shared with unauthorized users, or kept after the licence ends, which can breach contract terms and expose the organization to claims.

Who or what is affected: Analysts, managers and the data supplier

Causes

Broad sharing habits, unclear licence scope and renewals handled late.

Consequences

Contract claims, suspended supply of data, financial penalties under the contract and loss of supplier trust.

Existing controls

  • Licence terms are filed in a shared contract register.
  • Analysts sign an acceptable use statement for external datasets.

Control gaps

Licence terms are not linked to the datasets themselves, so usage is checked only from memory.

Additional controls

  • Prevent: Tag each licensed dataset with its permitted uses and block exports that exceed them.
  • Detect: Review dataset access logs against licence terms each quarter and record any exceptions.
Likelihood 4 x impact 4 before; 2 x 3 after Owner: Procurement and contracts lead Target: Within 45 days Early warning: A licensed dataset found in an extract shared outside the approved group.
H02

Sampling bias and poor source data quality

Operational Before 16 After 6

Dashboards and models are built on samples or source extracts that leave out groups, carry duplicate records or use stale codes, so reported trends mislead the people who rely on them.

Who or what is affected: Business managers and customers affected by their decisions

Causes

Convenience samples, changing source systems and missing reference data.

Consequences

Wrong targets, unfair treatment of some groups and loss of trust in the reports.

Existing controls

  • Source owners confirm row counts at each load.
  • A monthly data quality report lists null and duplicate rates.

Control gaps

No one checks whether a sample actually matches the population it claims to describe.

Additional controls

  • Prevent: Define the sampling frame and any weighting rules before analysis is published.
  • Detect: Compare sample profiles with the full population each month and flag gaps above agreed limits.
Likelihood 4 x impact 4 before; 2 x 3 after Owner: Data quality lead Target: Within 60 days Early warning: A segment whose record count falls sharply after a source system change.
H03

Unlawful use of personal data and consent gaps

Legal and compliance Before 20 After 8

Personal records are combined, profiled or shared for purposes that people did not agree to or that lack a lawful basis, which exposes individuals and the organization to enforcement and complaints.

Who or what is affected: Customers, employees and the privacy officer

Causes

Broad extracts, unclear purpose statements and consent records kept apart from the data itself.

Consequences

Regulatory action, complaints, compensation claims and lasting damage to customer trust.

Existing controls

  • Personal fields are listed in the data catalogue.
  • Analysts complete a privacy briefing before gaining access.

Control gaps

Consent status is not carried with the data when it moves into the warehouse.

Additional controls

  • Prevent: Pseudonymize identifiers in analytical sets and remove fields not needed for the stated purpose.
  • Respond: Stop the affected extract, notify the privacy officer and follow the agreed breach procedure.
Likelihood 4 x impact 5 before; 2 x 4 after Owner: Privacy officer Target: Within 30 days Early warning: A personal field appearing in a dashboard that was approved without it.
H04

Excess access rights and data kept beyond need

Technology and data Before 12 After 6

Users keep access after changing roles, and old extracts are kept on shared drives without a deletion schedule, so sensitive data is exposed and retention duties are not met.

Who or what is affected: Staff, customers and data stewards

Causes

Role changes without removal of old rights, copies made for quick tasks and no deletion schedule.

Consequences

Unauthorized disclosure, breach of retention limits and a larger impact when a breach does occur.

Existing controls

  • Access is granted through named role groups.
  • Warehouse snapshots are held for a fixed period.

Control gaps

Extracts saved to personal drives do not appear in any inventory.

Additional controls

  • Detect: Review access lists each quarter and remove rights that no longer match a current role.
  • Recover: Delete expired extracts on schedule and record each deletion for audit purposes.
Likelihood 3 x impact 4 before; 2 x 3 after Owner: Data steward Target: Within 45 days Early warning: Extract files older than the retention period found on a shared drive.
H05

Misleading dashboards and biased analytical models

Operational Before 12 After 6

Formulas, joins or model features are wrong or encode bias, and results reach decision makers without independent checks, so choices are made on figures that are incorrect or unfair.

Who or what is affected: Decision makers, customers and the analytics team

Causes

Complex joins, rushed releases and features that act as proxies for protected characteristics.

Consequences

Poor decisions, unfair outcomes, rework and reputational harm to the organization.

Existing controls

  • Calculated measures are documented in a metric dictionary.
  • Dashboards are peer reviewed before release.

Control gaps

Peer reviews check layout and labels but not whether the logic matches the metric definition.

Additional controls

  • Prevent: Require a test set with known answers for every new calculation and every model.
  • Detect: Monitor model outputs by group and compare them with prior runs after each change.
Likelihood 4 x impact 3 before; 2 x 3 after Owner: Analytics manager Target: Within 60 days Early warning: A metric whose value changes without a recorded change request.
H06

Pipeline failure and stale reporting

Technology and data Before 12 After 4

Scheduled loads, connectors or instrumentation fail silently or only in part, so reports show old or incomplete data while still appearing current to their users.

Who or what is affected: Report users, finance teams and operations teams

Causes

Source schema changes, expired credentials and network outages between systems.

Consequences

Decisions made on stale figures, missed reporting deadlines and costly rework.

Existing controls

  • The scheduler sends a failure alert to the on-call analyst.
  • Row counts are logged for each load.

Control gaps

Silent partial loads do not trigger an alert because no minimum row threshold is set.

Additional controls

  • Respond: Run a fallback load and mark affected dashboards as stale until the data is confirmed.
  • Recover: Rebuild failed loads from source snapshots and confirm totals against control figures.
Likelihood 3 x impact 4 before; 2 x 2 after Owner: Data engineering lead Target: Within 30 days Early warning: A load that completes with fewer rows than the trailing average.
H07

Irreproducible analysis and weak methodology

Operational Before 9 After 4

Results cannot be repeated because code, data versions, parameters or assumptions were not recorded, so reports cannot be checked or defended when their findings are questioned.

Who or what is affected: Analysts, reviewers and decision makers

Causes

Ad hoc work in spreadsheets, no record of working environments and staff turnover.

Consequences

Unverifiable claims, lost trust and wasted effort whenever results must be reproduced.

Existing controls

  • Analysis code is stored in a shared repository.
  • Key assumptions are written in each project summary.

Control gaps

Spreadsheet-based calculations sit outside version control and are not logged.

Additional controls

  • Prevent: Version datasets and parameters alongside the code for every published output.
  • Detect: Rerun a sample of reports each quarter from recorded inputs and compare the results.
Likelihood 3 x impact 3 before; 2 x 2 after Owner: Analytics manager Target: Within 90 days Early warning: A published figure that cannot be traced to a stored input version.
H08

Misuse of insights and missing ethical review

Reputational Before 8 After 3

Analyses that affect people, such as pricing, staffing or eligibility, go live without ethical review or a clear limit on use, so findings are applied in ways their evidence does not support.

Who or what is affected: Customers, employees and the wider public

Causes

Pressure to act quickly, unclear limits on use and no review step for people-related analysis.

Consequences

Unfair decisions, complaints, reputational damage and a loss of trust among staff.

Existing controls

  • Projects above a set impact level need sponsor approval.
  • Reports carry a stated intended use.

Control gaps

Intended-use statements are not checked against how the reports are actually used.

Additional controls

  • Prevent: Require an ethics review for any analysis that scores or ranks individuals before release.
  • Respond: Withdraw the output, record the misuse and inform the sponsor and the affected team.
Likelihood 2 x impact 4 before; 1 x 3 after Owner: Governance committee chair Target: Within 90 days Early warning: A report used for a purpose not listed in its intended-use statement.

Rating method and evaluation

Each risk is rated on two five-point scales. Likelihood runs from Rare to Almost certain, and impact runs from Insignificant to Severe. Multiplying the two gives a score from 1 to 25, which falls into one of four bands: Low from 1 to 4, Medium from 5 to 9, High from 10 to 16 and Critical from 17 to 25. The score is a guide for ordering work, not a precise measurement. Inherent risk is the score before any controls are applied. Residual risk is the score after the existing controls and the additional controls are in place, and it must never be higher than the inherent score. Impact is judged on harm to individuals, legal exposure, contract loss, wrong decisions and damage to trust, and a risk is never rated down only because it has not yet occurred. Anything still High needs senior attention and a dated plan, anything Medium needs routine controls with a named owner, and anything Low is accepted and watched. Analysts, stewards and the privacy officer should challenge the ratings, and every rating is reviewed when an indicator reaches its threshold or after a major change to a source or tool.

Likelihood, impact and rating bands

Likelihood

  • 5 Almost certain: Expected to happen in most circumstances.
  • 4 Likely: Will probably happen in many circumstances.
  • 3 Possible: Could happen at some time.
  • 2 Unlikely: Not expected, but it could happen.
  • 1 Rare: Only in exceptional circumstances.

Impact

  • 1 Insignificant: No lasting effect. Handled within routine work.
  • 2 Minor: Small effect that is quickly corrected.
  • 3 Moderate: Noticeable effect that needs managed effort to correct.
  • 4 Major: Serious effect on people, delivery or reputation.
  • 5 Catastrophic: Severe or lasting effect that threatens the whole activity.

Bands (likelihood x impact)

  • Low (1 to 4): Accept and monitor.
  • Medium (5 to 9): Manage with routine controls.
  • High (10 to 16): Senior attention and a treatment plan.
  • Critical (17 to 25): Act first. Do not proceed until reduced.

Monitoring, escalation and review

Review this assessment every 6 months, and sooner when an indicator below reaches its threshold.

Early-warning indicators and escalation thresholds
Indicator to watchEscalate when
Personal fields present in dashboards or extracts that were not approved for that use.Escalate to the privacy officer at once and suspend the output until the field is removed or approved.
Licensed datasets found in exports outside the approved group or permitted purpose.Escalate to the contracts lead within one working day, and to the head of analytics if the export has left the organization.
Scheduled loads completing with fewer rows than the trailing average by more than an agreed margin.Escalate to the data engineering lead immediately, and mark dependent dashboards as stale until confirmed.
Published figures that cannot be traced to a stored input version and parameter set.Escalate to the analytics manager within three days, and to the governance committee if the figure is used in a decision.
Extract files or access rights older than the retention period or outside current role needs.Escalate to the data steward within one week, and to the privacy officer if personal data is involved.

Contingency arrangements

If a serious event occurs, such as a personal data exposure, a supplier claim about licence misuse or a published figure found to be wrong, the head of analytics suspends the affected outputs and informs the governance committee. The data engineering lead freezes the related loads and restores data from the last verified snapshot. The privacy officer and contracts lead decide what must be reported and to whom, using the procedures for the relevant jurisdiction. Affected dashboards are marked as withdrawn until corrected, and managers are told plainly what changed and what to use in the meantime.

Prioritized action plan

  1. Restrict personal fields in analytical extracts and confirm a lawful basis for each, owner privacy officer, within 30 days.
  2. Tag every licensed dataset with its permitted uses and remove exports that exceed them, owner procurement and contracts lead, within 45 days.
  3. Set minimum row count thresholds on every scheduled load and alert on silent partial loads, owner data engineering lead, within 30 days.
  4. Define sampling frames and test a sample against its population for the top three reports, owner data quality lead, within 60 days.
  5. Require version records and an ethics review for people-related analysis, owner analytics manager and governance committee chair, within 90 days.

Conclusion

Trusted insight rests on a few steady habits: tagged licences, minimal personal data, checked samples, tested logic, recorded versions and clear limits on use. The highest scores sit with privacy and licensing, and each has controls that bring the residual score down. Treat the register as a living record, review it on the stated cycle, and seek qualified legal and privacy advice for any real question about personal data, contracts or regulated use.

Assumptions and limitations

Assumptions

  • The organization operates under one data protection regime, which is not named here, and holds a documented lawful basis for each personal dataset.
  • Licensed datasets come with written terms that limit purpose, sharing and retention.
  • Analysts use a shared warehouse and a version control system for code, though spreadsheets are still used in places.

Limitations

  • The scores are examples for learning and do not describe any real organization, supplier or dataset.
  • Privacy, licensing and records rules vary by jurisdiction and change over time, so specialist legal review is needed.
  • This is educational content and does not replace a formal site-specific assessment or licensed advice.

Related frameworks to study

  • Data governance operating model with named owners and stewards
  • Three lines model of risk ownership
  • Data protection principles of purpose limitation and minimization
  • Hierarchy of controls

Technical Risk Assessment

China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.

For Learning Reference Only: This assessment demonstrates how risk assessment principles are applied in a practical setting, providing a useful reference for learning and understanding. Actual assessments may be adapted to suit specific site conditions and circumstances.

Keep learning

Study the subject behind this assessment.

Business Intelligence Management course Read the best practices for this course

More Technical risk assessments

Hourly Active: 35,661 Learners