Employee onboarding: three signals to check in the first 90 days, not a checklist
Why employee onboarding stops after week one, what Bauer, Kammeyer-Mueller and Gallup found, and three checks for a new hire's first 90 days.
Employee onboarding is the period in which a new hire comes to know what is expected of them, believes they can deliver it and is accepted by the team. Most programs we see measure something else: the laptop handed over, the accounts created. Talya Bauer and colleagues' meta-analysis puts exactly those three signals at the centre.
A new hire's first day almost always goes well. The laptop is on the desk, the logins work, the team takes them to lunch. By Friday, every line on the onboarding checklist is ticked.
Week two looks different. The manager has a delivery due, the colleague assigned to help has deadlines of their own, and the new person reads documentation and waits to be given something to do. They do not ask much. A month in, nobody knows exactly how they are doing, because nobody has checked.
Put simply, employee onboarding usually ends where the task list ends, not where the person starts working in the role. The good news is that the gap shows up with three plain questions, asked by the manager on fixed dates across the first 90 days.
TL;DR
- Employee onboarding is finished when the new hire knows what is expected, believes they can deliver and is accepted by the team, not when the administrative checklist is complete.
- In a 2007 meta-analysis of 70 newcomer samples, Talya Bauer and colleagues found general support for a model in which role clarity, self-efficacy and social acceptance link socialization practices to outcomes such as job satisfaction, performance and intention to remain.
- The three adjustment checks: Role, Confidence, Acceptance. We ask them as three direct questions at days 7, 30, 60 and 90, and the manager asks them, not HR.
- John Kammeyer-Mueller and colleagues followed 264 newcomers through 14 waves of measurement and found that support from coworkers and supervisors declined within the first 90 days.
- Only 12% of US employees say their company does a good job of onboarding, according to Gallup's Emily Wetherell and Bailey Nelson (2021).
- The checklist stays with HR; the signals belong to the manager. Splitting those two responsibilities is the first change we make to an onboarding process.
What is employee onboarding, and where does it usually stop?
It is the process by which an organization helps a new hire become effective in the role and in the team. In practice, most processes stop at the administrative layer: contract, equipment, access, introductions. The hard part, learning what actually matters and who to work with, is left to chance.
The onboarding plans we see at clients are usually the same document: thirty lines, split by day, each owned by HR, IT or facilities. It is a good document, and it gets the person into the building and into the systems.
The trouble is what it leaves out. The checklist records that the person received the job description, not that they understand what matters in the role. It records that they were introduced to the team, not that the team has taken them in. It records that they finished the e-learning modules, not that they feel able to work on their own.
We are not the only ones to see the gap. Emily Wetherell and Bailey Nelson of Gallup report that only 12% of US employees say their company does a good job of onboarding (Wetherell and Nelson, Gallup, 2021). The figure describes American employees, and the article does not give a sample size.
The practical distinction is simple: onboarding tasks get ticked, adjustment signals can only be checked by asking. The first belongs to a process, the second to a conversation.
What does the research say about how new hires settle in?
Two things a manager can use. Bauer's meta-analysis puts three signals at the centre of adjustment: role clarity, self-efficacy and social acceptance. Kammeyer-Mueller's study shows that the support a new hire receives declines within the first 90 days, and that early support related more clearly to later outcomes than changes along the way did.
Three signals, not thirty tasks. Talya Bauer, Todd Bodner, Berrin Erdogan, David Truxillo and Jennifer Tucker tested a model of newcomer adjustment across 70 unique samples of new employees (Bauer, Bodner, Erdogan, Truxillo and Tucker, Journal of Applied Psychology, 2007). In the model, what the organization does (socialization tactics) and what the newcomer does (seeking information) feed three forms of adjustment: clarity about the role, confidence in being able to do the work, and acceptance by colleagues. Those three in turn connect to job satisfaction, organizational commitment, performance, intention to remain and actual turnover. The authors report that the results generally supported the model.
It is a meta-analysis that pools mostly correlational studies, so it does not prove that any particular onboarding practice produces retention. What it gives us is a map of the three things worth checking.
Support fades exactly when it matters. John Kammeyer-Mueller, Connie Wanberg, Alex Rubenstein and Zhaoli Song followed 264 newcomers through 14 waves of measurement and found that support from coworkers and supervisors declined within the first 90 days of employment (Kammeyer-Mueller, Wanberg, Rubenstein and Song, Academy of Management Journal, 2013). Early support and undermining had more significant relationships with outcomes measured after day 90 than did increases or decreases along the way. And undermining by the supervisor was uniquely associated with a higher risk of leaving.
None of these three sources comes from Romania: the Gallup figure describes US employees, and both academic studies were published by researchers working outside the country. We read the same direction in both studies, though, and it is an awkward one for the way onboarding is usually run: what matters happens in the relationship with the manager and the team, not on the checklist.
Why does employee onboarding stop after the first week?
Because the checklist has a clear owner, HR, and the adjustment signals have none. After week one the administrative tasks are done, so the process looks finished. The manager assumes HR has it covered, HR assumes the manager does, and the new hire says nothing, because they have only just arrived.
From what we see at clients, the stall has three mechanisms. Usually one dominates.
The wrong owner. Onboarding is written and tracked by HR, because HR did the hiring. But two of the three signals, role and acceptance, play out almost entirely between the new hire, their manager and the team. HR can build the frame. It cannot know, from a distance, whether the person has understood what matters by week five.
A buddy without a mandate. The colleague assigned to help usually gets the role with no time set aside and no idea what to watch for. They help when asked, and the new hire rarely asks.
The new hire's silence. In our experience, someone still in their probation period seldom volunteers that they are unsure what is expected of them. So the problem stays invisible until it becomes a performance problem. We have also found that "do you need anything?" opens nothing. A precise question does.
All three have the same fix: the signals need an owner, a date in the calendar and questions written in advance.
What are the three onboarding checks for a new employee?
Role: the new hire can say, in their own words, what matters in the job and how it is judged. Confidence: they believe they can do the work alone, or know exactly what they are missing. Acceptance: they have colleagues they will take a question to without hesitating. The manager checks all three at days 7, 30, 60 and 90.
We built them on the three forms of adjustment in Bauer's meta-analysis and phrased them as questions, because a question has an answer you can hear. The dates of 7, 30, 60 and 90 days are our design choice, not a research finding: we want a checkpoint before early support thins out, as Kammeyer-Mueller and colleagues observed it does.
1. Role
The question: "If you had to explain to a friend what matters most in your job, what would you say?"
If the person describes tasks rather than priorities, or puts first something that is fifth on your list, you have a clarity problem, not a competence problem. We wrote about why clarity is a property of the system, not of the documentation, in the Clarity capability.
What to change: on day 7, write three lines together: what matters in the role, what a good result looks like at day 90, who decides what. Reread them at every check.
2. Confidence
The question: "What part of your work would you not yet do on your own?"
The question assumes there is such a part, which makes it easy to answer. A new hire who says "nothing" on day 30 is either exceptional or does not yet feel safe admitting what they do not know. In our experience the second is more common, and it tells you something about acceptance as well.
What to change: for whatever they name, set a first real task with low stakes that they do alone the following week, with a short review afterwards.
3. Acceptance
The question: "Who on the team do you go to when you get stuck?"
If the answer is only "you", or only "my buddy", the person is connected to a single point, not to the team. When that point goes on holiday, adjustment stops.
What to change: give the new hire a working reason to go to two or three different colleagues. A coffee introduction is a fine start; we prefer to add a real work question as well, because it gives them a reason to come back.
How to use them together
The three checks fit into a fifteen-minute conversation, which can simply be your weekly one-to-one. We wrote about how to keep that meeting the employee's rather than the manager's in one-on-one meetings. On day 90, if all three have clear answers, onboarding is done. If not, it is not done, whatever the checklist says.
What does repaired onboarding look like?
The checklist stays with HR and finishes in week one. The signals move to the manager, who has four short checks in the calendar and three questions written in advance. The buddy gets protected time and knows what to watch for. The new hire learns on day one that they will be asked, which makes the questions easier to answer.
What follows is a composite of a pattern we see often, not any single client.
A services firm hires four people in one quarter, across two teams. Onboarding is a well-made checklist, and each person completes it in week one. Three months later, managers describe two of them as "still slow", and one has already started looking for another job.
The changes are small. HR keeps the checklist and adds a single new line: "adjustment checks are in the manager's calendar". Managers get the three questions on one page. Each buddy gets two hours a week blocked in the calendar for the first six weeks.
At the first 30-day check, a pattern appears that was invisible before: one of the "slow" hires was working correctly, but on the wrong priority, because the job description and what the team actually needed did not say the same thing. It was not a competence problem. It was a role problem, and a fifteen-minute conversation fixed it.
The three checks apply in the same way to a newly promoted manager: the desk stays the same, the role changes completely.
What you can do next
Take the most recent hire on your team, anyone from the last three months, and ask them the three questions this week. Not as an appraisal, but as a check: tell them in advance that you will do this with every new hire from now on.
If an answer surprises you, you have found exactly what the onboarding checklist had no way of showing. And if you are wondering what happens when these signals go unchecked for months, we wrote about that in why good employees leave.
If you want the checks to become a habit for every manager, not only the ones who read articles, Manager OS works on all four Capability OS domains, Clarity, Accountability, Dialogue and Learning, on your teams' real situations.
So we will leave you with a question. If you asked your most recent hire today what matters most in their job, would they say the same thing you would?
FAQ
What is employee onboarding?
It is the process by which a new hire becomes effective in the role and in the team. It includes the administrative layer but does not end there. It ends when the person knows what is expected, believes they can deliver and is accepted by colleagues: the three forms of adjustment in Bauer and colleagues' meta-analysis.
How long should onboarding last?
The administrative part usually takes a week. Adjustment takes longer: we work to a 90-day horizon, because that is the window Kammeyer-Mueller's study followed. Wetherell and Nelson write that, in Gallup's experience, new employees typically take around 12 months to reach their peak performance potential.
Who owns onboarding, HR or the manager?
Both, for different things. HR owns the process and the administrative checklist. The manager owns the signals: role clarity, the new hire's confidence in their own ability, and acceptance by the team. Problems start when each assumes the other has all of it covered.
What should you ask a new employee after 30 days?
Three precise questions rather than "how is it going?". What matters most in your job? What part of your work would you not yet do on your own? Who on the team do you go to when you get stuck? The answers show where adjustment stands on each of the three signals.
Sources
- Bauer, T. N., Bodner, T., Erdogan, B., Truxillo, D. M. and Tucker, J. S. (2007). Newcomer adjustment during organizational socialization: A meta-analytic review of antecedents, outcomes, and methods. Journal of Applied Psychology, 92(3), 707-721. Meta-analysis, 70 newcomer samples. The model with role clarity, self-efficacy and social acceptance linking socialization practices to outcomes was generally supported. Pools mostly correlational studies.
- Kammeyer-Mueller, J. D., Wanberg, C. R., Rubenstein, A. L. and Song, Z. (2013). Support, undermining, and newcomer socialization: Fitting in during the first 90 days. Academy of Management Journal, 56(4), 1104-1124. Longitudinal study, 14 waves, 264 newcomers. Support declined within the first 90 days; supervisor undermining was associated with a higher risk of leaving.
- Wetherell, E. and Nelson, B. (2021). 8 Practical Tips for Leaders for a Better Onboarding Process. Gallup. US data: only 12% of employees say their company does a good job of onboarding. No sample size given. Also the source for the estimate, presented as Gallup's experience, of around 12 months to peak performance potential.
By Szilárd Kacsó