Supporting Graduate Mental Health at Work: A Manager’s Guide to the First 90 Days

Why the transition deserves deliberate design

Young workers are not a single or fragile group. They bring different resources, responsibilities and experiences. Research nevertheless identifies psychosocial work factors such as job quality, control, uncertainty and respect as relevant areas, while also warning that the evidence specific to young workers is heterogeneous and often low-certainty.

That uncertainty is a reason for careful local listening, not inaction. The first 90 days contain predictable demands: unfamiliar standards, ambiguous social rules, frequent evaluation and a high need for feedback. An organisation can reduce unnecessary uncertainty without removing useful challenge.

Evidence: 1

Build role clarity in layers

A job description is not role clarity. New starters need examples of good output, an explanation of how priorities are chosen and permission to ask when two expectations conflict. Break the role into learnable stages and state what competence should look like at 30, 60 and 90 days.

Formal onboarding evidence is limited, but a systematic review found the strongest support for structured, supported on-the-job training. Pair information with practice: observe a task, complete it with support, receive specific feedback and repeat with increasing independence.

Evidence: 2

Make support a system, not a personality test

Do not make a new starter guess who is safe to ask. Give them a manager for priorities, a peer for everyday questions and a clear route for sensitive or formal concerns. Explain confidentiality and the limits of each role.

Schedule check-ins before problems appear. A weekly 20-minute conversation can cover what is clear, what is confusing, where demand feels high and which decision is needed. This supports help-seeking without requiring the graduate to present a crisis before receiving attention.

Evidence: 3

Train managers to respond—not diagnose

WHO recommends manager training to improve knowledge and attitudes, help managers recognise and respond to employees who need support and identify job stressors. It is not training to diagnose or provide treatment.

A useful response is calm and concrete: listen, thank the person for telling you, ask what support would help, explain what you can change and signpost qualified help. Record only what is necessary under the organisation’s policy. If there is an immediate safety concern, follow the appropriate emergency process.

Evidence: 3

Teach mental fitness in context

Early-career resilience training should connect directly to real moments: receiving corrective feedback, managing competing deadlines, recovering after a presentation, asking for clarification and separating one difficult outcome from a judgement about personal worth.

Avoid implying that a breathing technique will fix an unmanageable placement or unsafe team. Pair skills with manager commitments around workload, feedback quality, inclusion and recovery. The message should be “you have tools and we will improve the conditions,” not “use these tools to tolerate anything.”

Evidence: 3, 4

A first-90-days framework

In days 1–30, prioritise belonging, practical orientation and clear first tasks. In days 31–60, increase ownership while keeping feedback frequent and specific. In days 61–90, review workload, strengths, development goals and the support network—not only completed outputs.

Ask the graduate what would have made the transition clearer and safer, then use that learning to improve the next cohort. Aggregate themes so onboarding evolves without exposing an individual’s private concerns.

• Days 1–30: people, purpose, tools, expectations and small successful repetitions.
• Days 31–60: broader tasks, explicit decision rights and two-way feedback.
• Days 61–90: sustainable workload, development plan and support review.
• Throughout: protected questions, regular manager contact and clear routes to help.

Confidence grows faster when asking for help is treated as part of learning, not evidence that learning has failed.

Evidence: 2, 3