RAID Logs, Issues and Dependencies: Keeping an IT Project Unblocked

When a project is running, problems and blockers arrive daily: a server is not ready, a decision is overdue, a team member is pulled onto another job. The difference between smooth and chaotic projects is rarely the number of problems but how they are tracked and resolved. A RAID log gives the project manager one place to track the four things that most often derail delivery: Risks, Assumptions, Issues and Dependencies. Some teams add Decisions (RAIDD) or Actions.

The Four Elements

ElementDefinitionExample
RiskAn uncertain future event that could affect objectivesThe data centre migration may overrun the weekend window
AssumptionSomething taken to be true without proof; if wrong, it creates riskThe finance team will provide test data by 10 February
IssueA problem that is happening now and needs resolvingThe test environment is down and testing cannot proceed
DependencyA relationship where one piece of work relies on another, often outside your controlOur go-live depends on the network team completing the firewall changes

The elements interact. An unvalidated assumption is a latent risk; a risk that occurs becomes an issue; an overdue dependency becomes an issue. Review them together so that connections are visible. Risk is covered in depth in Project Risk Management.

Issue Management

An issue management process turns problems into owned, time-bound actions:

  1. Capture the issue with a clear description, date raised, who raised it and effect on the project.
  2. Assess its severity and urgency, for example by effect on critical path, cost or quality.
  3. Assign an owner and a target resolution date.
  4. Resolve: investigate options, decide and act.
  5. Escalate if it cannot be resolved within the owner’s authority or by the target date.
  6. Close it with a record of the resolution, and review patterns for lessons.

Issue Log Template

IDIssueRaisedSeverityImpactOwnerActionDueStatus
I-014UAT environment unavailable; database refresh failed3 MarHighUAT start at risk; critical pathInfrastructure leadRestore from last backup; add monitoring5 MarIn progress
I-015Vendor invoice dispute over extra days4 MarMediumPayment delay; relationship strainProject managerReview change log with vendor lead8 MarOpen

Managing Dependencies

Dependencies are among the biggest causes of IT project delay, because they sit in other teams’ priorities. Types include:

  • Internal: between work packages within your project.
  • External: on vendors, regulators, customers or other projects.
  • Technical: on systems, interfaces, environments or data.
  • Resource: on shared people or equipment.

Practical habits: list dependencies during planning and add them to the schedule as milestones; agree dates and owners with the providing team in writing; check progress before the due date, not on it; draw a simple dependency map for complex programmes; and create a fallback if the date is at risk. For Agile teams, tools such as a program board visualise cross-team dependencies and make them visible to everyone.

Assumptions: The Hidden Risk

Every plan rests on assumptions, such as “the API will be available”, “two testers will be assigned” or “the legacy data is clean”. The danger is that assumptions go unrecorded and unchallenged. For each, note the owner who will validate it and the date by which it must be confirmed. When an assumption proves false, convert it into a risk or issue immediately and update the plan.

Escalation

Escalation means raising a matter to a higher level of authority because it exceeds your authority or cannot be resolved at the current level. It is a normal management tool, not an admission of failure. Define escalation paths and thresholds in the communication plan, for example:

LevelWhoTypical trigger
1Team lead or project managerIssue within the team’s control
2Functional manager or vendor managerUnresolved after 2 working days, or resource conflict
3Project sponsorThreatens milestone, budget variance above threshold, or needs a policy decision
4Steering committee or executiveThreatens project viability or needs cross-organisation decision

How to escalate well

  1. Escalate early, while there are still options.
  2. State the problem and its impact in one or two sentences.
  3. Describe what you have already tried.
  4. Offer options with a recommendation and the decision you need.
  5. Give a deadline for the decision and the cost of delay.
  6. Follow up and record the outcome.

Compare these two messages. Weak: “We have a problem with the environment.” Strong: “UAT cannot start on Monday because the test database is corrupted. If it is not restored by Wednesday, go-live moves two weeks. Infrastructure can restore it if we get priority from the platform manager. Can you ask him to approve this today?”

Decisions and Action Logs

Many project disputes stem from forgotten decisions. A short decision log records what was decided, by whom, when, why and what alternatives were rejected. An action log records tasks arising from meetings, with owner and due date. Review overdue actions at each meeting, because unowned actions are quietly dropped.

Running the RAID Review

Hold a short weekly review with the core team and update the log before status reporting. Ask for each item: Is the owner still right? Has the date or rating changed? Is escalation needed? Are there new items? Report the top items to stakeholders, and archive closed entries so the log stays readable. Tools range from a shared spreadsheet to features in Jira or project management software; see Project Management Tools Compared. For Agile teams, see Impediment Management in Scrum.

Common Mistakes

  • Keeping the log but never reviewing it.
  • Entries with no owner or date.
  • Mixing up risks and issues, or leaving assumptions unrecorded.
  • Waiting too long to escalate, then presenting a crisis instead of a choice.
  • Tracking dependencies only inside your own team.
Key takeaways
  • A RAID log tracks risks, assumptions, issues and dependencies in one place.
  • Every entry needs an owner, a date and a next action.
  • Validate assumptions; manage external dependencies actively and in writing.
  • Escalate early, with impact, options, a recommendation and a deadline.

Related Lessons