Risk, Change and Performance Analytics

Introduction

Construction and site management is conducted in an environment of uncertainty. Design information may be incomplete, materials may be delayed, labour productivity may vary, regulations may change, and client decisions may alter the scope after work has begun. In interior projects, these uncertainties are intensified by tight programmes, specialist trades, coordination of services, procurement of finishes and the need to protect occupied buildings.

Risk management, change control and performance analytics provide an integrated management system for dealing with this uncertainty. Risk management anticipates what may happen; change control governs what is actually altered; performance analytics shows whether the project is achieving its intended results. Used together, they support timely, evidence-based decisions rather than reactive crisis management.

This chapter examines methods suitable for advanced diploma-level practice, including risk registers, qualitative and quantitative assessment, change-control workflows, key performance indicators (KPIs), productivity measurement, trend analysis and dashboard design.

Learning objectives

By the end of this chapter, the learner should be able to:

1. Explain the relationship between risk, issue, change and performance management.
2. Create and maintain a project risk register.
3. Assess probability, impact, exposure, proximity and risk response effectiveness.
4. Design a controlled workflow for variations and other changes.
5. Evaluate time, cost, quality, safety and stakeholder consequences of change.
6. Calculate and interpret selected construction performance measures.
7. Use trends, thresholds and dashboards to support corrective action.
8. Recognise ethical, legal, technological and governance requirements in project analytics.

1. The management logic: uncertainty, change and control

A risk is an uncertain event or condition that, if it occurs, may have a positive or negative effect on project objectives. A negative risk is commonly called a threat; a positive risk is an opportunity. An issue is a condition that has already occurred and requires action. A change is an approved or proposed alteration to scope, design, sequence, resources, specifications, contract terms or completion requirements.

These concepts should not be confused:

A sound control system maintains traceability: the original risk, the decision taken, the responsible person, evidence of implementation and the resulting effect should be connected in project records.

2. Risk management framework

Risk management is a continuous cycle rather than a one-time workshop. A practical sequence is:

1. Establish context and objectives.
2. Identify risks.
3. Analyse probability and impact.
4. Evaluate priorities against agreed thresholds.
5. Plan responses.
6. Assign ownership.
7. Monitor, review and communicate.

The project context should define objectives, stakeholders, interfaces, assumptions, constraints, risk appetite and escalation rules. For an interior fit-out, relevant categories may include design coordination, procurement, labour, logistics, existing-building conditions, statutory compliance, health and safety, quality, cost, programme, sustainability, security and reputational risk.

2.1 Risk identification

Risk identification should use several sources rather than relying only on individual judgement. Useful techniques include design reviews, site inspections, constructability reviews, lessons-learned records, supplier consultations, programme analysis, interviews, checklists, assumption testing and workshops with trade contractors.

A well-written risk statement has three parts:

Cause → uncertain event → effect

For example: “Because the existing ceiling drawings may be inaccurate, concealed services may conflict with the proposed partition layout, causing redesign, delay and additional cost.” This formulation distinguishes the source of uncertainty from the event and its consequences.

Risk descriptions should be specific, evidence-based and linked to an objective. “Poor coordination” is weak; “unresolved mechanical and lighting coordination above the reception ceiling may require rework before handover” is more actionable.

2.2 The risk register

The risk register is the principal working record. A typical register includes:

A register is not effective if it is merely produced for an approval gate. It should be reviewed at design coordination meetings, procurement reviews, site progress meetings and formal reporting periods. Closed risks should remain archived so that the project retains its learning history.

2.3 Qualitative assessment

A common screening method assigns probability and impact scores, often from 1 to 5. The simple risk priority score is:

where R is the qualitative risk score, P is the probability rating and I is the impact rating. Both ratings are dimensionless ordinal scores, not measured percentages. Therefore, the result ranks attention; it does not represent a monetary loss or a precise probability.

A risk matrix may classify scores such as 1–4 as low, 5–9 as moderate, 10–16 as high and 17–25 as extreme. These bands are organisational conventions and must be defined in the project risk plan. A matrix should assess impact separately for cost, time, quality, safety, compliance and reputation where a single composite score would conceal an important consequence.

For example, a safety risk with moderate cost impact may still require immediate action because safety consequences have a different tolerance threshold. The highest material consequence should not be hidden by averaging several categories.

2.4 Quantitative exposure

When adequate data exist, expected monetary value (EMV) can support contingency analysis:

where pᵢ is the probability of scenario i, expressed between 0 and 1, and cᵢ is its financial consequence in currency units. The formula assumes that scenarios are sufficiently defined and that the expected value is appropriate for the decision. It does not predict the exact outcome of an individual project.

For a single threat with a 0.25 probability of causing a £12,000 loss:

If three mutually exclusive scenarios have probabilities 0.50, 0.30 and 0.20, with consequences of £0, £8,000 and £20,000 respectively, the expected cost is:

Probabilities should sum to 1 for mutually exclusive and collectively exhaustive scenarios. If events are correlated, simply adding separate EMVs may understate or overstate exposure. For major projects, simulation, sensitivity analysis or expert-supported probabilistic modelling may be more appropriate.

For schedule exposure, a three-point estimate may be used where optimistic, most-likely and pessimistic durations are available. The PERT expected duration is:

where a is the optimistic duration, m the most-likely duration and b the pessimistic duration, all in the same time unit. This estimate assumes a beta-like distribution and is a planning approximation, not a guarantee. If a = 4 days, m = 7 days and b = 16 days:

2.5 Response strategies

For threats, common responses are:

• Avoid: change the design, method or sequence so the threat no longer exists.
• Mitigate: reduce probability or impact, for example by early surveys or mock-ups.
• Transfer: allocate financial or operational responsibility through insurance, warranty or contract, while recognising that accountability cannot always be transferred completely.
• Accept: take no preventive action beyond monitoring, with an agreed contingency.
• Escalate: refer a risk beyond the project manager’s authority to the client, board or programme manager.

For opportunities, responses include exploit, enhance, share, accept and escalate. Each response should specify an action, owner, due date, required budget and success measure. A response that is not resourced is an intention, not a control.

3. Change control in construction and interior projects

Change is not inherently harmful. Design development, statutory requirements, value engineering and client improvement may create beneficial changes. Uncontrolled change, however, produces scope ambiguity, claims, rework, procurement disruption and unsafe working conditions.

3.1 Sources and types of change

Typical sources include client instructions, design errors or omissions, unforeseen existing conditions, statutory changes, substitutions, value engineering, specialist contractor proposals, access restrictions, supply disruption and corrective work. Changes may affect:

• scope and quantities;
• materials, finishes and performance specifications;
• programme and sequence;
• labour, plant and logistics;
• contract price and cash flow;
• quality, maintainability and warranties;
• health, safety and environmental controls.

A change may be a variation under the contract, a design development, a non-conformance correction or an instruction requiring formal authorisation. The contract’s definition and notice requirements govern the legal treatment.

3.2 A controlled workflow

A robust workflow is:

1. Initiate: record the request, source, reason and required decision date.
2. Screen: determine whether it is a genuine change, a clarification, a defect correction or an existing obligation.
3. Define: prepare revised drawings, specifications, quantities and assumptions.
4. Assess: evaluate cost, time, quality, safety, compliance, procurement, carbon and operational effects.
5. Consult: obtain input from designers, commercial staff, site management, specialist trades and the client.
6. Recommend: state accept, reject, defer or request more information.
7. Authorise: obtain approval from the person with contractual and financial authority.
8. Implement: issue controlled information, update programme, budget, risk register and method statements.
9. Verify: inspect the completed change and confirm testing, commissioning and records.
10. Close: reconcile cost and time, archive evidence and record lessons learned.

No physical work should normally proceed on a cost-bearing change without appropriate written authority, except where immediate action is necessary to protect life, property or the works. Emergency instructions should be documented retrospectively as soon as practicable.

3.3 Impact assessment

Change assessment should show both direct and indirect effects. Direct cost may include labour, materials, plant and subcontractor charges. Indirect cost may include supervision, temporary works, storage, remobilisation, disruption, preliminaries, design fees and financing effects.

Time analysis should identify affected activities, logic relationships, available float and the contractual completion date. A change to a non-critical activity may have no completion-date effect, while a short activity on the critical path can create substantial exposure. Claims should be supported by contemporaneous programmes, instructions, records and causation analysis rather than by a simple assertion that “the change caused delay.”

Safety assessment must consider altered access, temporary stability, dust, noise, hot works, manual handling, hazardous materials, isolation of services and interaction with occupants. The change should trigger review of risk assessments and method statements where the work method or hazard profile changes.

3.4 Change log and configuration control

The change log records request number, description, originator, status, estimate, decision, approver, implementation date and links to documents. Configuration control ensures that only the current approved drawing, specification and schedule are used. Revision clouds, transmittals, document registers and digital permissions reduce the risk of construction from obsolete information.

4. Performance measurement and analytics

Performance measurement converts project records into evidence for action. A KPI is useful only when it has a clear definition, owner, data source, frequency, target, tolerance and response. A dashboard that displays many numbers without decision rules is a reporting display, not a management system.

4.1 Balanced KPI design

A balanced project scorecard should include leading and lagging indicators across multiple dimensions.

A KPI should not encourage undesirable behaviour. For example, maximising installation quantity without measuring defects may reward rushed work. Balanced indicators reduce this risk.

4.2 Core construction calculations

Productivity is commonly expressed as output per labour input:

where Q is accepted output, such as square metres installed, and H is labour-hours. Units must be explicit, for example m²/labour-hour. Some organisations instead report labour-hours per unit:

These are reciprocals when both use the same output and labour definition. Output should be accepted and defect-free where possible; otherwise the metric may reward rework.

If a team installs 180 m² of flooring using 72 labour-hours, productivity is:

and unit labour is 0.40 labour-hours/m². Comparisons require similar conditions, including substrate readiness, material type, access, crew composition and measurement boundaries.

Planned progress can be compared with actual progress using a simple schedule performance ratio:

A ratio below 1 indicates less physical progress than planned for the chosen period. It does not, by itself, explain the cause or prove critical-path delay.

Earned value analysis provides a more structured cost and schedule view. The principal terms are:

• PV (planned value): authorised budgeted value of work scheduled by the status date.
• EV (earned value): authorised budgeted value of work actually completed by the status date.
• AC (actual cost): actual cost incurred for completed work by the status date.

The schedule variance and cost variance are:

Both are in currency units and depend on the project’s approved performance measurement method. Positive SV means more budgeted work has been completed than planned; positive CV means the earned value exceeds actual cost. They should not be interpreted without considering the baseline and the quality of reported progress.

The schedule performance index and cost performance index are:

These ratios are dimensionless. Values above 1 are favourable relative to the baseline; values below 1 are unfavourable. They are indicators, not forecasts of completion by themselves.

If PV = £100,000, EV = £90,000 and AC = £95,000:

The project is behind the baseline in earned progress and is obtaining approximately £0.947 of budgeted value per £1 of actual cost. Management must still investigate whether the causes are design information, access, procurement, rework or measurement bias.

A basic estimate at completion can be calculated as:

where BAC is the budget at completion. This form assumes that the observed cost efficiency will continue and that remaining work will be performed according to the original budget basis. Alternative EAC methods are appropriate when past variance is exceptional or when remaining work has been re-estimated. It should therefore be labelled as a forecast assumption, not a certainty.

4.3 Data quality and trend analysis

Trend analysis examines a measure over time rather than relying on one reporting period. A seven-period moving average, for example, is:

where xₜ is the current observation and the terms are the six preceding observations. This smooths short-term fluctuation but introduces delay and should not replace investigation of sudden deterioration. The periods must be comparable and the measure consistently defined.

Control charts may distinguish common-cause variation, inherent in a process, from special-cause variation caused by an unusual event. However, control limits should be based on adequate, stable data; arbitrary limits can create false alarms. A trend becomes managerial evidence when it is connected to cause, threshold and action—for example, three consecutive weeks of declining first-pass quality triggering a review of substrate preparation and trade supervision.

Data quality should be checked for completeness, accuracy, timeliness, consistency, validity and uniqueness. Digital dashboards can make unreliable data appear authoritative. The project should identify the system of record, measurement owner, calculation definition and audit trail for every critical KPI.

5. Dashboards for corrective decision-making

A useful dashboard has three layers:

1. Outcome layer: overall time, cost, quality, safety and client status.
2. Diagnostic layer: trends, exceptions, risk heat maps, change ageing and productivity by work package.
3. Action layer: owner, due date, decision required, escalation status and expected effect.

Colour coding should be governed by thresholds. Green, amber and red must mean defined conditions, not personal opinion. A dashboard should display the reporting date, baseline version, data freshness and exclusions. It should separate actuals from forecasts and approved changes from pending requests.

Corrective action follows a cycle: detect deviation, diagnose cause, select intervention, assign owner, set due date, implement, verify effect and close. If productivity falls, adding labour may not solve the problem if the real constraint is late design information or restricted access. Analytics should therefore support root-cause analysis, not merely trigger resource increases.

6. Integrated case study: illustrative interior fit-out

Background

An illustrative project involves the refurbishment of a 2,400 m² corporate workplace in an occupied building. The scope includes partitions, ceilings, lighting, flooring, joinery and audiovisual systems. The planned completion period is 20 weeks. The client requests a revised acoustic wall finish during week 6.

Problem

The proposed finish has a 10-week procurement period rather than the 6 weeks assumed in the procurement schedule. At the same time, weekly first-pass quality performance for partitions has fallen from 92% to 76%, and actual installation productivity has declined from 2.4 to 1.8 m² per labour-hour. The change may affect cost, programme, quality and safety because additional access and working at height are required.

Analysis

The project team records the procurement threat in the risk register: inaccurate lead-time assumption may delay the acoustic finish, affecting handover and client approval. Probability is rated 4/5 and programme impact 4/5, producing a qualitative score of 16 under the project matrix. The risk is high, but the score does not replace an impact assessment.

The change request is screened as a client-initiated scope change. The designer confirms the technical performance requirement; the supplier provides a quotation and delivery commitment; the planner identifies affected activities; the site manager reviews access, storage and installation risks; and the commercial manager assesses direct and indirect cost.

The team considers three options:

The client approves the equivalent finish subject to a physical sample and performance confirmation. This is not an automatic substitution: the specification, fire performance, acoustic evidence, colour and warranty are verified before approval.

The productivity decline is investigated rather than attributed immediately to labour. The team finds that workfaces are being released late because electrical containment is incomplete and that repeated partition inspections are identifying avoidable alignment defects. The corrective plan includes a daily workface-readiness check, coordination of containment before partition closure, a sample panel, first-off inspection and revised short-interval planning.

Outcome

The approved change is entered in the change log, the risk is reduced to residual probability 2/5 and impact 3/5, and the programme baseline is updated only after formal approval. The equivalent finish arrives in week 11. Productivity recovers to 2.3 m² per labour-hour over the next three weeks, while first-pass quality rises to 90%. The dashboard reports the remaining procurement risk as amber because supplier delivery is still being monitored.

Learning points

The case demonstrates that risk, change and performance data are interdependent. A change assessment without productivity and quality evidence would have been incomplete. A dashboard showing only cost might have concealed the workface constraint. Formal approval, current information, measurable controls and verified outcomes are all necessary for reliable management.

7. Governance, ethics and professional responsibility

Risk and performance information influences payment, claims, employment decisions and client confidence. Deliberately delaying an adverse report, manipulating progress quantities or excluding defects from productivity data is professionally unacceptable. Managers should distinguish fact, estimate, assumption and judgement.

Confidential commercial information should be access-controlled. Personal data from workforce monitoring must be collected lawfully, proportionately and transparently. Safety data should not be used to discourage incident reporting. A low recorded incident rate may indicate under-reporting rather than excellent safety.

Contractual notices, approvals, inspection records and changes should be retained in accordance with the contract, organisational policy and applicable law. Regulatory requirements concerning building safety, fire performance, occupational health, environmental protection and accessibility take precedence over dashboard targets or commercial pressure.

8. Technology, AI and contemporary practice

Common digital tools include common data environments, mobile inspection applications, model-based coordination, automated quantity extraction, scheduling systems, business-intelligence dashboards and sensor-based monitoring. Their value depends on consistent data structures, defined ownership and integration with contractual processes.

AI may assist with classifying risks, detecting unusual trends, summarising records or forecasting likely delays. It should not independently approve a variation, certify safety compliance or replace professional judgement. Training data may contain bias; generated outputs may be inaccurate or lack traceability. Human review, source verification, access control, version management and audit logs are essential.

Contemporary practice is moving toward integrated digital twins, near-real-time progress capture, predictive analytics and carbon-aware procurement. These developments can improve anticipation, but they also increase dependence on data quality and cyber-security. The fundamental management questions remain: Is the information reliable? What decision will it support? Who is accountable? What evidence will demonstrate that the decision worked?

9. Best-practice principles

Effective risk, change and performance management is characterised by the following principles:

• link every risk and change to a project objective;
• use clear causal descriptions and named owners;
• distinguish risk, issue, change, forecast and actual result;
• assess time, cost, quality, safety and compliance together;
• update baselines only through authorised governance;
• measure accepted output, not merely activity;
• combine leading indicators with lagging outcomes;
• investigate causes before prescribing corrective resources;
• show data definitions, dates, assumptions and uncertainty;
• preserve an auditable record of decisions and evidence;
• escalate safety and statutory concerns without commercial suppression;
• review whether responses reduced residual exposure;
• capture lessons while information is still available.

Chapter Summary

Risk management provides a structured method for identifying uncertainty, assessing probability and impact, selecting responses and monitoring residual exposure. A risk register becomes useful when it contains specific causal descriptions, owners, response actions, deadlines, controls and links to issues or changes. Qualitative scores such as R = P × I support prioritisation, while EMV and three-point estimates can support quantitative planning when assumptions and data are adequate.

Change control protects the project from unauthorised scope, design and information changes. A complete workflow initiates, defines, assesses, authorises, implements, verifies and closes each change. Impact assessment must cover cost, programme, quality, safety, procurement, compliance and operational consequences.

Performance analytics converts project records into management evidence. Productivity, progress ratios, earned value, trends and quality measures should be interpreted in context. KPIs require clear definitions, owners, targets, thresholds and corrective actions. Dashboards are effective when they reveal exceptions and assign decisions, not when they merely display attractive numbers.

Finally, reliable management depends on ethical reporting, contractual traceability, data governance and professional judgement. Digital tools and AI can strengthen prediction and coordination, but they cannot remove accountability for safe, lawful and evidence-based decisions.

Key Terms

Risk: An uncertain event or condition that may affect project objectives.

Issue: A condition or event that has already occurred and requires action.

Change: An approved or proposed alteration to project scope, design, resources or requirements.

Risk register: The controlled record of identified risks, assessments, owners, responses and status.

Risk owner: The person accountable for managing and reporting a specified risk.

Probability: The assessed likelihood that a risk event will occur.

Impact: The consequence of a risk or change on an objective.

Residual risk: Exposure remaining after controls or responses are applied.

Contingency: Time, money or resources reserved for identified uncertainty.

Expected monetary value: Probability-weighted financial consequence across defined scenarios.

Change log: A record of proposed, approved, rejected and implemented changes.

Configuration control: Management of approved versions of project information.

KPI: A defined indicator used to evaluate performance against an objective or target.

Productivity: Output achieved per unit of input, such as m² per labour-hour.

Earned value: Budgeted value of work actually completed.

Planned value: Budgeted value of work scheduled by a specified date.

Actual cost: Cost incurred for work performed by a specified date.

Trend analysis: Examination of performance observations over time to identify direction or change.

Dashboard: A visual management interface displaying status, trends, exceptions and actions.

Residual exposure: The remaining time, cost, quality, safety or other consequence after response measures.

Review Questions

Short Answer Questions

1. Distinguish a risk from an issue and give one construction example of each.
2. What information should be recorded in a risk register?
3. State two purposes of configuration control.
4. Define a leading KPI and provide one interior-project example.
5. Why should productivity be measured using accepted output where possible?

Long Answer Questions

1. Explain the complete risk-management cycle for an interior fit-out project.
2. Describe a change-control workflow and explain why implementation should normally follow formal authorisation.
3. Compare qualitative risk scoring with expected monetary value, including limitations of each method.
4. Explain how cost, schedule, quality and safety KPIs can be combined in a balanced dashboard.
5. Discuss the ethical and governance responsibilities associated with construction performance data.

Analytical / Application Questions

1. A flooring team installs 240 m² using 96 labour-hours. Calculate productivity and unit labour, state the units, and explain two conditions that must be controlled before comparing this result with another team.
2. A project has PV = £150,000, EV = £135,000 and AC = £144,000. Calculate SV, CV, SPI and CPI, then interpret the results and identify two possible causes requiring investigation.
3. A client change has a 0.30 probability of causing a £18,000 loss and a 0.10 probability of causing a £40,000 loss; assume the scenarios are mutually exclusive and all other outcomes have zero loss. Calculate the expected monetary value and explain why the result should not be treated as a guaranteed cost.

Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes a risk from an issue?

A. A risk has already occurred, while an issue may occur later.

B. A risk is uncertain, while an issue has already occurred.

C. A risk is always financial, while an issue is always technical.

D. A risk requires no owner, while an issue requires an owner.

Correct Answer: B. A risk is uncertain, while an issue has already occurred.

Question 2

A risk has a probability rating of 4 and an impact rating of 3 on a defined five-point matrix. What is its qualitative score?

A. 7

B. 12

C. 43

D. 0.75

Correct Answer: B. 12

Question 3

A threat has a probability of 0.25 and a financial consequence of £12,000. What is its expected monetary value?

A. £3,000

B. £12,250

C. £48,000

D. £2,500

Correct Answer: A. £3,000

Question 4

Which action should normally occur before implementing a cost-bearing change?

A. Delete the original drawing immediately.

B. Obtain appropriate formal authorisation.

C. Update the dashboard but not the contract records.

D. Ask the trade contractor to proceed verbally.

Correct Answer: B. Obtain appropriate formal authorisation.

Question 5

A team installs 180 m² using 72 labour-hours. What is its productivity?

A. 0.40 m²/labour-hour

B. 2.00 m²/labour-hour

C. 2.50 m²/labour-hour

D. 252 m²/labour-hour

Correct Answer: C. 2.50 m²/labour-hour

Question 6

If PV is £100,000 and EV is £90,000, what is the schedule performance index?

A. 0.90

B. 1.11

C. £10,000

D. £190,000

Correct Answer: A. 0.90

Question 7

Which is a leading safety indicator?

A. Number of injuries recorded last year

B. Total final accident cost

C. Open high-risk safety actions past their due dates

D. Historical lost-time injury frequency

Correct Answer: C. Open high-risk safety actions past their due dates

Question 8

What is a principal limitation of a simple qualitative risk score P × I?

A. It cannot rank risks at all.

B. It produces a precise monetary forecast.

C. It uses ordinal ratings and does not represent exact probability or loss.

D. It can only be used for safety risks.

Correct Answer: C. It uses ordinal ratings and does not represent exact probability or loss.

Question 9

What is the primary purpose of a dashboard action layer?

A. To replace all project records.

B. To display decorative graphics.

C. To identify owners, due dates and decisions required.

D. To conceal unfavourable trends.

Correct Answer: C. To identify owners, due dates and decisions required.

Question 10

A project has EV = £90,000 and AC = £95,000. What does a negative cost variance indicate under EV terminology?

A. The earned value exceeds actual cost.

B. Actual cost exceeds earned value.

C. The project is ahead of schedule.

D. No cost information is available.

Correct Answer: B. Actual cost exceeds earned value.