An onboarding plan defines which knowledge and tasks a new hire should master by which point, who supports them, and how completion is recognised. What separates it from a welcome checklist is that it ties goals to dates, ownership and a verifiable outcome.
The practical value doesn’t come from the document; it comes from someone tracking it. That’s where most plans fail.
What belongs in an onboarding plan
Four entries per line — no fewer, and no more needed:
Goal
What should the person be able to do by the end of this stage? Not “get to know the system”, but “can create and release an order independently”.
Content
What actually happens — which conversation, which training, which task, which document.
Ownership
Who supports this step? Without a name, nothing happens.
Evidence
How is completion recognised? A signature, a finished task, a passed record, a conversation.
Plans without the last two columns are wish lists. They read well and stop being opened after two weeks.
Template: onboarding plan in five phases
This structure adapts to most roles. The timings are guidance — in regulated environments much of it moves earlier, because instruction has to be complete before independent work begins. Open the phase you need.
| Goal | Content | Owner | Evidence |
|---|---|---|---|
| Able to work from day one | Access, workspace, devices, software accounts | IT / manager | Access tested |
| Arrival clarified | Time, place, contact, plan for day one | Manager | Message sent |
| Mandatory set known | Which instruction the role requires | Manager | List available |
| Goal | Content | Owner | Evidence |
|---|---|---|---|
| Orientation | Tour, team, safety instruction | Manager | Instruction documented |
| Role understood | Tasks, expectations, responsibilities | Manager | Conversation held |
| First point of contact clear | Who answers which questions | Buddy / mentor | Introduced |
| Goal | Content | Owner | Evidence |
|---|---|---|---|
| Core processes seen | The three to five workflows the role touches daily | Buddy / mentor | Documentation read |
| First own task | A clearly bounded task with feedback | Buddy / mentor | Task completed |
| Mandatory training started | Instruction required before independent work | Manager | Attendance recorded |
| Goal | Content | Owner | Evidence |
|---|---|---|---|
| Core tasks independent | Main duties without constant checking back | Buddy / mentor | Observation confirmed |
| Interfaces known | Collaboration with adjacent areas | Manager | Conversations held |
| Interim review | What’s working, what’s missing, what wasn’t reached | Manager | Conversation documented |
| Goal | Content | Owner | Evidence |
|---|---|---|---|
| Role profile met | Comparison against the position’s requirements | Manager | Target-actual compared |
| Remaining records | All required instruction completed | Manager | Records complete |
| Development | What comes next over the following months | Manager | Goals agreed |
Creating an onboarding plan: the approach
Start from the role’s target profile, not the person.
Which knowledge and authorisations does this position require? The plan is the distance from what someone brings to what the role needs. Without a defined target there’s no measure — which makes the skills matrix the sensible basis.
Schedule mandatory training first.
Anything required before independent work belongs at the beginning, not the end. Which instruction applies per role and how renewal cycles run is covered on the training management software page.
Three to five goals per phase, not twenty.
Plans fail from overloading more often than from gaps. What isn’t achievable in week one belongs in month one.
A name against every goal.
Without named ownership the step gets dropped — particularly when the supporting person is busy with their own work.
Build evidence in from the start.
What has to be documented anyway should appear in the plan as evidence, rather than being assembled separately later.
Revise after first use.
The first plan for a role is a draft. Whatever the first new hire didn’t understand is missing from it.
Why templates alone don’t solve the problem
A spreadsheet template is a good start and entirely sufficient for a first hire. What it can’t do shows from the third onwards:
It doesn’t know whether something was completed. A tick in a cell is a claim. Whether the instruction actually took place and a record exists sits somewhere else.
It doesn’t know the role’s requirements. When the mandatory set changes, existing copies of the template don’t change with it.
It ends with onboarding. What comes afterwards — renewal deadlines, expiring records, new requirements — appears in no onboarding plan.
It belongs to one person. When the manager changes, the state of onboarding is often untraceable.
Moving from template to system isn’t a convenience question; it’s the point where a document becomes a traceable process.
Who makes sure the plan is actually worked through
The effort isn’t in writing the plan, it’s in staying with it — through weeks when the manager has other things to do.
So in Kollegia Antoine, one of the Digital Colleagues, takes that on: it derives from the role profile which mandatory training is outstanding for a newly filled position, flags when an agreed step hasn’t been completed by its date, and reminds about the interim review in time. Once onboarding ends, the same monitoring continues — for renewal deadlines and expiring records.
Three principles
- Permission intersection — a digital colleague never sees more than the person it works for. Onboarding data is personnel data; access follows existing permissions exactly.
- Fail-closed — when in doubt, nothing happens. If a detail is unclear, the process stops rather than guessing.
- Audit trail — every notification and change is logged and traceable.
FAQ
Source: Section 12 Arbeitsschutzgesetz (ArbSchG) – Unterweisung, German Federal Ministry of Justice.
Four entries per step: the goal (what should the person be able to do), the content (what happens), the ownership (who supports it) and the evidence (how completion is recognised). Plans without ownership and evidence tend not to get tracked.
Three months is common, structured as day one, week one, month one and month three. What matters more than total duration is the point at which the person takes on core tasks independently — usually the end of the first month.
Onboarding describes the whole integration process, including organisational and social aspects. The onboarding plan is the professional part of it: which knowledge is built by when.
For the first few hires, yes. Beyond several role profiles and mandatory trainings with deadlines it reaches its limits: it doesn’t show whether a step was actually completed, doesn’t update when requirements change, and ends with onboarding while renewal deadlines continue.
The responsible manager, based on the role’s requirement profile. It helps to involve the person doing the job today — they know most precisely what’s actually needed in the first weeks.
The sequence changes: required instruction has to be completed and documented before any independent work. Evidence there isn’t a by-product, it’s the precondition for the next step.
