Project Risk Management: Identify, Analyse, Respond and Monitor (With a Risk Register Example)
Every project contains uncertainty. A key developer may resign, a vendor may deliver late, a regulation may change, or a new technology may work better than expected. Risk management is the discipline of dealing with that uncertainty deliberately instead of reacting to surprises. A risk is an uncertain event or condition that, if it occurs, has a positive or negative effect on one or more project objectives. Negative risks are threats; positive ones are opportunities.
Risk vs Issue
A risk has not happened yet; an issue is a problem that is happening now. When a risk occurs, it becomes an issue and moves from the risk process to issue management; see RAID Logs, Issues and Dependencies. Writing “Server crashed” in the risk register is a category error; “Server may crash under peak load” is a risk.
The Risk Management Process
- Plan risk management: define approach, roles, categories, scales and thresholds.
- Identify risks: find and document them.
- Perform qualitative analysis: rank risks by probability and impact.
- Perform quantitative analysis: numerically assess the effect of the most significant risks (when justified).
- Plan responses: decide what to do about each important risk.
- Implement responses: carry out the agreed actions.
- Monitor risks: track, review, close old ones and find new ones.
Identifying Risks
Use many sources and techniques:
- Brainstorming workshops with the team and stakeholders.
- Checklists based on past projects; for IT: security, integration, performance, data quality, vendor delivery, skills availability, requirement volatility.
- Review of assumptions and constraints in the charter and scope statement.
- WBS review: ask of each work package, “What could go wrong?”
- Interviews with experienced people and subject experts.
- SWOT analysis (strengths, weaknesses, opportunities, threats).
- Root cause analysis and a risk breakdown structure to categorise by source.
Write each risk in the cause-event-effect format: “Because the payment vendor has not confirmed the API specification (cause), the integration design may be wrong (event), which could add three weeks to the schedule (effect).” This makes risks specific and actionable.
Qualitative Analysis: Probability-Impact Matrix
Rate each risk on probability and impact using simple scales, such as 1 to 5, and multiply to get a score.
| Impact → / Probability ↓ | 1 Very low | 2 Low | 3 Medium | 4 High | 5 Very high |
|---|---|---|---|---|---|
| 5 Very high | 5 | 10 | 15 | 20 | 25 |
| 4 High | 4 | 8 | 12 | 16 | 20 |
| 3 Medium | 3 | 6 | 9 | 12 | 15 |
| 2 Low | 2 | 4 | 6 | 8 | 10 |
| 1 Very low | 1 | 2 | 3 | 4 | 5 |
A common rule: scores of 15 and above are red (act now), 8 to 14 amber (manage actively) and below 8 green (monitor). Define the scales precisely in advance. For example, impact 4 might mean a delay of 3 to 6 weeks or a cost overrun above 10 per cent. Also assess proximity (how soon it might occur) and detectability.
Quantitative Analysis: Expected Monetary Value
Expected monetary value (EMV) = probability × impact in money. Consider three risks:
| Risk | Probability | Impact (₹ lakh) | EMV (₹ lakh) |
|---|---|---|---|
| Vendor API delivered late | 40% | −8 | −3.2 |
| Key developer resigns | 20% | −10 | −2.0 |
| Reusable component saves effort (opportunity) | 30% | +5 | +1.5 |
| Total expected effect | −3.7 |
The total EMV of −₹3.7 lakh provides a rational basis for sizing the contingency reserve (see Project Budgeting and Cost Control). Other techniques include decision trees, sensitivity analysis (tornado diagrams) and Monte Carlo simulation, which runs thousands of scenarios to show the probability of meeting a date or budget. Quantitative analysis is worth the effort only on larger, higher-stakes projects with decent data.
Risk Response Strategies
| Strategy | For threats | For opportunities | Example |
|---|---|---|---|
| Escalate | Pass to a higher authority if outside your control | Same | Raise a cross-project vendor risk to the programme board |
| Avoid | Remove the threat, often by changing the plan | Exploit: ensure it happens | Drop an unproven technology and use a mature one |
| Mitigate | Reduce probability or impact | Enhance: raise probability or benefit | Build a prototype early to test the integration |
| Transfer | Shift impact to a third party | Share: partner to capture the benefit | Fixed-price contract; insurance; SLA penalties |
| Accept | Take no action (passive) or hold contingency (active) | Accept the benefit if it arises | Hold a reserve for minor delay |
Choose responses whose cost is proportionate to the risk exposure. Also consider secondary risks (new risks created by the response) and residual risks (what remains after the response). Assign every response an owner and a due date, and define triggers, early warning signs that a risk is about to occur, plus a contingency plan (“plan B”) for if it does.
Risk Register Template
| ID | Risk (cause-event-effect) | P | I | Score | Response | Owner | Trigger | Status |
|---|---|---|---|---|---|---|---|---|
| R-007 | Because the vendor has not confirmed the API spec, the integration design may be wrong, adding 3 weeks | 4 | 4 | 16 | Mitigate: prototype the API calls in sprint 1; agree spec sign-off date in contract | Tech lead | Spec not received by 15 Dec | Open |
| R-011 | Because only one engineer knows the legacy billing system, absence could stall migration | 3 | 4 | 12 | Mitigate: pair a second engineer; document interfaces | PM | Engineer on leave notice | Open |
Monitoring Risks
Review the register in every status meeting, not just at kick-off. Ask: Has any probability or impact changed? Have triggers appeared? Are responses working? Are there new risks? Close risks that can no longer occur and record those that materialised. Report the top five or ten to the sponsor with a trend arrow. Agile teams do similar work continuously; see Risk Management in Scrum.
Risk Appetite and Culture
Organisations differ in how much risk they will accept for the potential reward. A bank may have low appetite for security risk but high appetite for experimental products. Know the sponsor’s tolerance and the thresholds that trigger escalation. Also build a culture where people raise risks without blame; a risk register full of bad news is a sign of health, whereas an empty one suggests that nobody is looking.
Common Mistakes
- Creating a risk register at kick-off and never updating it.
- Listing vague risks such as “poor communication” with no cause or effect.
- Confusing risks and issues.
- Identifying risks but assigning no owners or actions.
- Focusing only on threats and ignoring opportunities.
- A risk is an uncertain future event; an issue is a problem happening now.
- Write risks as cause, event and effect; score them by probability and impact.
- EMV = probability × impact and helps size contingency.
- Five response strategies: escalate, avoid, mitigate, transfer and accept (opportunities: exploit, enhance, share).
- Review risks continuously, with owners, triggers and plan B.