# The skills matrix: map to make reliable. Who can do what, on which instrument, with what level of autonomy? In the quality control laboratory, this question often goes unanswered. The **skills matrix** answers it — and reveals where the gaps that generate errors and OOS are hiding. Who · What · Autonomy A QC laboratory relies on dozens of techniques and systems — HPLC, GC, Karl Fischer, UV, LIMS… The matrix cross-references each one with each technician. Field experience — pharmaceutical quality control laboratory in France. The **skills matrix** is a simple but powerful table: technicians in rows; techniques and systems in columns; a status at each intersection. It turns a vague feeling ("the team is skilled") into an **objective map** of strengths and weaknesses — the prerequisite for managing reliability seriously. "Trained" does not mean "autonomous." A checked-off training record does not guarantee real mastery of a technique. Without a clear distinction between _not trained_, _trained but not autonomous_ and _trained & autonomous_, critical analyses get handed to people who are still learning — and the problem is discovered after the fact, through an OOS. The matrix makes this nuance visible, workstation by workstation. ## A level of mastery, not a checked box. Status 1 ### Not trained The technician does not work on the technique. Useful for sizing the training need and the dependency risk. Status 2 ### Trained, not autonomous Still learning: they operate under mentoring or double-check. The riskiest status if mistaken for autonomy. Status 3 ### Trained & autonomous Qualified, they operate alone and can mentor others in turn. The goal to reach, critical technique by critical technique. ## Four readings, four decisions. ### Fragile techniques A column full of "not trained" flags a technique that depends on one or two people: a continuity and error risk. ### Exposed people A row light on autonomy — often a recent hire — points to high exposure to error and OOS. ### The "key-person" risk A critical technique mastered by a single expert: what happens if they leave or fall ill? ### The upskilling plan The matrix becomes an action plan: who to train, on what, in what order — prioritized by criticality. ## From the map to the learning path. A matrix is only worth something if it triggers action. Every "not trained" or "not autonomous" cell on a critical technique calls for a **qualification path**: concepts, mentored practice, supervised routine, then formal validation of autonomy. The matrix steers, the path trains, qualification proves — and analyst error goes down. [Build the path with 70-20-10 →](/en/blog/blended-learning-70-20-10-gmp/) ## Skill, reliability, compliance. [Quality control ### OOS & onboarding Why the root cause of an OOS often lies in onboarding training. Read the article →](/en/blog/oos-human-error-onboarding-qc-lab/) [Platform ### Manage skills Learning Suite: learning paths, qualifications and skills traceability, audit-ready. See the platform →](/en/learning-suite/) [Consulting ### Document & capture Capture critical know-how and make it transmissible, before it walks out the door. Learn more →](/en/consulting/documentation-training/) ## The skills matrix in practice. What is a skills matrix in the QC laboratory? + It's a table that cross-references technicians (in rows) and analytical techniques or systems (in columns), with a mastery status at each intersection: not trained, trained but not autonomous, trained and autonomous. It objectifies the team's real skills. Why distinguish "trained" from "autonomous"? + Because a recorded training does not prove mastery. Conflating the two means handing critical analyses to people who are still learning — a direct driver of error and OOS. The distinction lets you supervise (mentoring, double-check) until autonomy is qualified. How does the matrix help reduce errors? + It reveals fragile techniques, exposed people and "key-person" risks, and turns into an upskilling plan prioritized by criticality. By targeting training where the risk is real, it prevents analytical errors at the source. How do you keep it up to date? + By embedding it into day-to-day management: update it at every qualification, arrival or departure, and review it regularly. Paired with a skills-management platform, it stays alive and provides traceability that holds up in inspection. ## Know who can do what — and make it reliable. Let's map your laboratory's skills and build the upskilling plan.