Skip to content

06 · Risk Management Basics

A risk is an uncertain event or condition that, if it occurs, has an effect — positive or negative — on at least one project objective (scope, schedule, cost, or quality). Most of the working vocabulary treats "risk" as synonymous with "threat," but the formal definition includes opportunities too (a positive risk — e.g., "a new library ships early and could cut our development time") even though threats dominate day-to-day attention. Risk management is not about eliminating uncertainty — that's impossible on any project worth doing — it's about identifying it early, deciding deliberately what to do about each risk, and tracking it so nothing important is forgotten. This module covers the foundational tools: identification, the probability/impact grid, and the RAID log. A deeper, more quantitative treatment (qualitative and quantitative risk analysis) is in Level 2.

Identifying risks

Risk identification works best as a structured, repeatable exercise rather than a one-time brainstorm at kickoff — new risks surface throughout a project's life as more becomes known. Common techniques:

Technique How it works
Brainstorming Team session listing anything that could go wrong (or right)
Checklist review Compare against a standing list of risks common to this type of project
Assumption analysis Examine every assumption from the scope statement — if it's wrong, what happens?
Expert judgment Ask people who've run similar projects what bit them
SWOT analysis Strengths/Weaknesses/Opportunities/Threats, viewed from the project's angle

A well-written risk statement follows a simple structure: "Because of [cause], [risk event] may occur, which would lead to [effect]." For example: "Because the vendor has never integrated with our specific payment processor before, the integration may take longer than estimated, which would lead to a schedule slip on the launch date." This structure forces you past a vague worry ("the vendor integration is risky") into something specific enough to actually plan around.

Assessing risks: the probability/impact grid

Once risks are identified, they need to be prioritized — not every risk deserves the same attention. The standard tool is a probability/impact grid, plotting how likely a risk is against how severe its effect would be:

Low Impact Medium Impact High Impact
High Probability Medium priority High priority Critical priority
Medium Probability Low priority Medium priority High priority
Low Probability Low priority Low priority Medium priority

Placing each identified risk into this grid does two things: it tells you where to spend your limited planning time (Critical and High-priority risks need an active response plan; Low-priority risks may just need to be watched), and it gives you a defensible, repeatable way to explain to a sponsor why one risk is being actively managed and another is being accepted without action.

The four response strategies

For each risk that warrants active management, there are four standard response strategies (the same four apply to opportunities, framed positively — see the note below):

Strategy What it means Example
Avoid Change the plan to eliminate the risk entirely Choose a vendor with a proven integration instead of an unproven one
Mitigate Reduce the probability or impact Run an early technical spike with the vendor to surface integration problems before they're on the critical path
Transfer Shift the risk (or its financial consequence) to a third party Buy insurance, or write a fixed-price contract that puts overrun risk on the vendor
Accept Acknowledge the risk and do nothing proactive (usually reserved for low-priority risks) Note it in the log and revisit if it starts to materialize

The same four strategies apply to opportunities

For positive risks, the mirror strategies are Exploit (make it happen), Enhance (increase the probability/impact), Share (partner with a third party to help it happen), and Accept (welcome it if it comes, but don't chase it).

The RAID log

A RAID log is the single running document that ties risk management into the rest of project execution — it stands for Risks, Assumptions, Issues, and Dependencies, tracked together because they're closely related: an unmanaged assumption often becomes a risk, and an unmanaged risk that materializes becomes an issue.

Type Definition Example entry
Risk Something that might happen "Vendor integration may take longer than estimated"
Assumption Something being taken as true without proof "Assuming the vendor's sandbox environment is available for testing by week 2"
Issue Something that has already happened and needs resolving now "Vendor's sandbox environment is not yet available — blocking integration testing"
Dependency Something the project relies on from outside its own control "Legal sign-off on the vendor contract, owned by the Legal team"

Reviewing the RAID log on a fixed cadence (weekly, on most projects) is what keeps risk management from becoming a one-time kickoff exercise that gathers dust — the log should visibly change every review, with new entries added, statuses updated, and closed items moved to a resolved section rather than deleted (deleted history is lost learning).

A worked example: a risk that was identified but not actively managed, and what happened

A construction project's risk register flags, in week 1: "Because the site has known soil-stability issues from a prior geological survey, foundation work may require additional engineering, which would lead to a schedule delay." Probability is assessed as Medium, impact as High — landing in the High-priority cell of the grid. The response strategy chosen is Mitigate: commission an updated geotechnical survey before foundation work begins, at a cost of $8,000, specifically to reduce the probability of discovering the issue mid-excavation (which would be far more expensive to address once equipment is already mobilized).

Contrast this with a second, hypothetical version of the same project where the same risk was identified but assigned "Accept" by default, because no one wanted to spend the $8,000 up front. Three weeks into excavation, the soil issue does materialize — it has now become an Issue in the RAID log, not a risk — and remediation costs $45,000 and adds three weeks to the schedule, because it's being solved reactively with equipment already mobilized and idle, rather than proactively with a clean survey result in hand. The $8,000 mitigation, in the real version of this project, was cheap insurance against a downside that was over five times as expensive to absorb after the fact — which is the general pattern that makes proactive risk management worth the up-front time it costs.

How It Actually Works

A probability/impact matrix isn't just a visual sorting tool — the risk score it produces, Score = Probability × Impact, is a simplified Expected Monetary Value (EMV) calculation, the same math used quantitatively in Level 2. If probability and impact are each scored 1–5, a risk rated (4, 4) = 16 "should" outrank one rated (5, 2) = 10 on the heat map, but that ordinal multiplication silently distorts real dollar risk: a 20% chance of a $500,000 loss (EMV = $100,000) can rate lower on a 1–5 heat map than an 80% chance of a $50,000 loss (EMV = $40,000) if the ordinal buckets are drawn carelessly, even though the first is 2.5× worse in real expected-dollar terms. This is exactly why qualitative scoring is a triage step, not a final answer — it exists to cheaply rank dozens of risks so you know which handful deserve the more expensive quantitative EMV or Monte Carlo treatment (Level 2 Module 4, Level 3 Module 3). The RAID log's four categories map to four different control actions with different mechanics: Risks are managed by changing probability or impact before they occur, Issues (materialized risks) are managed by containment and recovery cost after the fact, which is why an issue's true cost is always closer to its full impact than its pre-event EMV — the probability term has collapsed to 1.

Exercise

For the project you've been building through this level's exercises, identify five risks using the "because of [cause], [risk event] may occur, which would lead to [effect]" structure. Place each on the probability/ impact grid and note its priority level. For the two highest-priority risks, write which of the four response strategies (avoid/mitigate/ transfer/accept) you'd choose and why. Then start a RAID log with at least one entry in each of the four categories (risk, assumption, issue, dependency) for your project.