DEV Community

keyur shah
keyur shah

Posted on

The RAID Log: One Document, Four Categories, No More Surprises

In any project, hidden pitfalls often surface at the worst possible moment. A well‑maintained RAID log—Risks, Assumptions, Issues, Dependencies—keeps those surprises in check. By treating the log as a living artifact rather than a kickoff checklist, teams gain visibility, accountability, and a clear path to mitigation.

Risks

Risks are “what‑might‑happen” items that have not yet materialised. Each risk should be recorded with an estimate of likelihood (e.g., low, medium, high) and potential impact (e.g., cost, schedule, quality). Crucially, a mitigation plan must accompany every risk; a risk without a response is just noise. For example, “Third‑party authentication service could experience downtime during peak load” would be logged with a 30 % likelihood, high impact, and a mitigation such as “Implement fallback token cache and schedule load‑testing by week 4.”

Assumptions

Assumptions are statements taken as true without verification—often hidden drivers of risk. An example might be “Vendor API will be production‑ready by June.” Because an unvalidated assumption can become a surprise later, the team should treat each assumption as a mini‑risk: validate it early, assign an owner, and record the validation date. If the API delivery slips, the assumption instantly transforms into a risk that can be mitigated.

Issues

Issues are risks that have already materialised. They demand immediate action, an owner, and a due date. For instance, “The UI prototype failed accessibility testing” becomes an issue with a clear owner (UX lead) and a resolution deadline (end of sprint 3). Logging an issue without follow‑up leads to “just another line item” that never gets resolved.

Dependencies

Dependencies are work items that the team relies on from external sources—another team, a vendor, or a shared platform. Each dependency should have a “needed‑by” date so that any delay becomes visible before it blocks the project. Example: “Data warehouse schema update required for reporting” with a needed‑by date of week 5. If the data team pushes the date, the log highlights the slip, prompting a schedule adjustment.

Do / Avoid

Do

  • Review the RAID log weekly, not just at kickoff.
  • Assign a clear owner to every item; ownership drives accountability.
  • Convert realized risks into issues immediately, updating status and due dates.

Avoid

  • Letting assumptions sit unvalidated for the entire project lifecycle.
  • Logging a risk without attaching a mitigation plan.
  • Treating the RAID log as a one‑time kickoff document; it must evolve.

A disciplined RAID log turns unknowns into known work items, keeping projects on track and surprises at bay.


Book a call

Top comments (0)