Small companies onboard badly for understandable reasons: nobody owns it, everyone is busy, and the new person is expected to work it out. The cost is a slower ramp, and sometimes a departure within a year that was entirely avoidable.
A structured process does not require an HR function. It requires someone to own it and a written plan.
Before day one
The most neglected phase and the cheapest to fix. Everything here should be done before they arrive.
- Equipment ordered and arriving before the start date, not on it.
- Accounts created for every system they will need, tested.
- A written 90-day plan sent in advance: what success looks like at 30, 60 and 90 days.
- The first week’s calendar, populated and sent.
- An announcement to the team saying who is joining, what they will do and why.
- A named buddy who is not their manager, briefed on what that involves.
A new starter whose laptop is not ready learns something about the company on their first morning, and it is not a good thing to learn.
Week one: belonging and context
The goal is not productivity. It is that they understand what the company does, who everyone is, and how work happens.
Day one should include a conversation with their manager about the role and the plan, a tour of the tools, and something concrete completed — however small. Ending day one having shipped a documentation fix or answered a support ticket is worth more than a day of presentations.
The rest of the week: short introductory conversations with the people they will work with, time with customers or customer recordings, and the context nobody writes down — why the product is the way it is, which decisions are settled and which are live.
Resist filling the week with meetings. Leave time to read, explore and be confused in peace.
The first 30 days: a real contribution
By the end of the first month they should have delivered something genuinely useful, independently.
Choose the first project carefully. It should be real rather than make-work, small enough to finish, low enough risk that a mistake is recoverable, and representative of the work they will actually do.
Shipping something real in the first month is the strongest predictor of how someone feels about a new job. Spending a month reading documentation produces a competent, disengaged employee.
Days 30 to 90: independence
The second and third months are about reducing supervision.
By day 60 they should be handling routine work without checking, contributing in planning, and able to find answers themselves. By day 90 they should be fully independent in the core of the role and starting to improve something rather than only operating it.
Hold a structured conversation at each milestone against the written plan. These are not performance reviews; they are a check that the plan was right and that the support is adequate.
What to write down
Documentation is what makes onboarding scale, and small companies always underestimate how much tacit knowledge they are carrying.
The highest-value items: how to set up a development environment or equivalent, how work moves from idea to delivery, who owns what, where decisions are recorded, and the glossary of internal terms — which is the one most often missing and most confusing to a newcomer.
A good trick: ask each new starter to improve the onboarding documentation as part of their first month. They are the only people who can see what is missing, and the window closes within weeks.
The manager’s part
Onboarding is largely a management responsibility and it is where managers under pressure cut first.
Three commitments worth protecting. A weekly one-to-one from week one, not from month two. Explicit feedback early, because small corrections at week three are easy and the same conversation at month four is difficult. And visible availability — a new starter who feels they are interrupting will stop asking, and that is when problems start accumulating quietly.
Remote onboarding
Everything above matters more when the person is not in a room with anyone.
Written documentation becomes essential rather than helpful. The buddy relationship matters more and should be with someone in a compatible time zone. Introductory conversations need scheduling deliberately because they will not happen by accident. And more frequent manager contact in the first fortnight compensates for the absence of ambient signal. See remote team management.
Measuring it
Four things worth tracking even informally.
- Time to first meaningful contribution. Should be days, not weeks.
- Time to full productivity, as judged by the manager against the 90-day plan.
- Retention at 12 months, which correlates strongly with the first-month experience.
- Feedback at 30 and 90 days. Ask what was confusing and what was missing, and act on it before the next hire.
The 30-day feedback is the most useful and the shortest-lived. After a couple of months people have normalised the gaps and can no longer see them.
Role-specific onboarding
The structure above is general. Three roles need specific additions that are routinely missed.
Engineers need a working environment on day one and a first change in production within the first week, however trivial. Shipping something small immediately teaches the whole delivery path and is the fastest way to feel competent. See CI/CD practices for why a fast, reliable pipeline makes this possible.
Salespeople need customer conversations early — listening to calls in week one, joining them in week two. Product training in isolation produces someone who can describe features and cannot have a discovery conversation.
Customer-facing support need to answer real tickets quickly, with review, rather than reading documentation for a fortnight. The documentation makes sense once there is a question attached to it.
The common thread is real work early, with support, rather than preparation in the abstract.
When onboarding is going wrong
Four signals, all visible within the first six weeks.
They are not asking questions. A new starter who asks nothing in week two is either exceptional or has concluded that asking is unwelcome. The second is far more common.
They are still waiting for access to something. Three weeks in, this says nobody owns the process.
They have not completed anything. Whatever the reason, it needs addressing now rather than at the first review.
They are working on something other than the plan. Usually because the team pulled them into whatever was urgent. Understandable, and it means the 90-day plan has quietly been abandoned.
Each of these is recoverable in week three and much harder in month four, which is the argument for the structured check-ins rather than waiting for a formal review.
Frequently asked questions
How long should onboarding last?
Ninety days as a structured process, with the intensive part in the first two weeks. Senior and specialist roles often take longer to reach full independence.
Who should own onboarding in a small company?
The hiring manager owns the role-specific plan; one person should own the general process and keep the documentation current. Leaving it entirely to individual managers produces wide variation.
What is the most common mistake?
Nothing being ready on day one. It is entirely preventable and it sets the tone for everything that follows.
Should new starters have goals in the first 90 days?
Yes, written and agreed in advance, framed as learning and contribution milestones rather than as performance targets.
