Your latest new hires are six to eight weeks into their jobs. The welcome survey came back positive, and nobody has resigned yet. So, is your onboarding working?
The honest answer is that you do not know yet, and the indicator you would trust most, retention after a year, will arrive long after you can do anything about it. This article explains why onboarding is so hard to measure, why day 60 is the right moment to look, which signals to track, and how to turn the results into action before day 90.
Why onboarding is so hard to measure
Every HR team knows the problem. The outcomes that matter are slow, and the signals that arrive quickly are not very reliable. Gallup observes that new employees typically take around 12 months to reach their peak performance potential. Meanwhile, the Work Institute 2025 Retention Report finds that as much as 40% of all turnover happens within the first year.
Put these together and the trap is clear. If you wait for attrition data to judge your onboarding, the cohort you are judging has already left. You learn something useful for the next cohort, but nothing that helps the people who just arrived.
Early satisfaction flatters the result
The fast signal most companies rely on is the welcome survey. The problem is timing. Research on the honeymoon-hangover effect, published in the Journal of Applied Psychology by Wendy Boswell, John Boudreau and Jaya Tichy, shows that job satisfaction rises sharply after a voluntary job change, then drifts back down. A survey sent in the first week mostly measures the honeymoon.
Newcomers are polite
There is also a human bias. Most newcomers are still on probation at day 60. They want to make a good impression, and they will rarely tell a survey that might reach their manager that they feel lost or alone. A high score from someone with no one to talk to is not a good sign. It is a missing signal.
Why day 60 is the right checkpoint
The research on newcomer adjustment points to a narrow window. A study of 264 newcomers followed over 14 waves, published in the Academy of Management Journal, found that support from coworkers and supervisors tends to decline over the first 90 days, and that the support received early on best predicts how newcomers are doing after three months.
Day 60 sits at the right point in that curve. The welcome glow has faded, so answers reflect daily reality. Patterns are visible: who is attending meetings, who has stopped. And there is still a month left before day 90 to fix what is broken. In many countries it is also the last useful moment before probation reviews, when a newcomer who feels isolated may simply decide not to stay.
Leading and lagging indicators: measure both, act on one
The way out of the measurement trap is to separate two kinds of indicators. Lagging indicators confirm what happened: early attrition, time to productivity, performance at one year. Leading indicators warn you of what is about to happen, while you can still change it.
Keep tracking the lagging indicators by cohort, as we suggested in our guide to the 4 Cs of onboarding. But at day 60, the leading ones are where your attention should go.
What to measure at day 60
A meta-analysis of 70 samples of newcomers by Talya Bauer and colleagues identifies three conditions that explain whether onboarding leads to satisfaction, commitment, performance and retention: role clarity, self-efficacy and social acceptance. They make a natural backbone for a day-60 check-in, with a fourth signal that most surveys miss: what newcomers actually do.
1. Role clarity: ask both sides
Role clarity is the strongest predictor in the research and the easiest to fix. The useful measure is not only the newcomer's answer, but the gap between their answer and their manager's.
- Ask the newcomer to rate "I know what is expected of me in my role" on a scale of 1 to 5.
- Ask the manager to rate how clear they think the role is for this person.
- Treat a large gap as a priority, whichever direction it goes.
2. Confidence: does the newcomer feel able to do the job?
Self-efficacy is the newcomer's belief that they can succeed in the role. It drops quietly when someone is struggling and has no one to ask, long before it shows in performance.
- Ask "I feel confident I can do my job well" on the same scale.
- Cross it with role clarity: low clarity and low confidence together usually mean the onboarding plan stopped after week one.
3. Social integration: the signal you can observe, not just ask about
Social acceptance is the condition most often left to chance, and the slowest to build. A Microsoft Research study of more than 10,000 new employees, presented at the ACM Web Conference in 2023, found that new hires' networks keep growing during onboarding but remain well below those of tenured colleagues even after six months. In BambooHR's 2023 onboarding survey, one new hire in five said their company did nothing at all to help them make friends at work.
The good news is that integration is the one condition you can measure through behavior as well as through a survey.
- Ask "I know who to ask for help outside my own team" and "I have at least one colleague I can talk to openly".
- Count the colleagues each newcomer has actually met outside their team since their start date.
- Check how many buddy meetings took place. Microsoft found that the value of a buddy grows with every meeting, as we detailed in our guide to buddy programs.
4. Participation: the early warning nobody looks at
Disengagement rarely announces itself. It shows up as a declined invitation, a skipped coffee, a buddy meeting that keeps getting postponed. A newcomer who stops showing up to the moments designed to connect them is sending a signal weeks before any survey picks it up. We made the same point about existing teams in why quiet quitting is a connection problem.
- Look at participation trends for your latest cohort of new hires, not just the overall rate.
- Flag newcomers whose participation has dropped since their first weeks.
- Cross that list with low integration scores: those are the people to reach out to first.
A day-60 scorecard you can run in one afternoon
None of this needs a new tool or a long survey. Five questions and three numbers are enough to see where your cohort stands.
A few rules make the results trustworthy:
- Keep answers confidential. Share results by cohort or team above a minimum group size, never individually with managers.
- Read the distribution, not just the average. An average of 4 can hide three people at 1.
- Use your own cohorts as a baseline. Each cohort's numbers become the benchmark for the next one, which is more relevant than any industry average.
Turn the results into action before day 90
A checkpoint is only worth running if it changes something. Here is what to do with the answers in the four weeks that remain.
- For individuals with low integration: a direct, informal conversation with HR, a new buddy if the first pairing did not work, and two or three introductions to colleagues in teams they work with.
- For teams where newcomers met no one outside the team: plan cross-team coffees for the whole cohort, as described in our guide to breaking down silos.
- For clarity gaps: ask the manager for a 30-minute role conversation before day 90, with the newcomer's answers as a starting point.
- For the whole cohort: share back what you heard and what will change. Newcomers who see their feedback used answer the next survey more honestly.
How RandomCoffee helps you see the part that is hardest to measure
Clarity and confidence are measured by asking. Integration and participation are measured by watching what happens, and that data usually sits nowhere. If you run onboarding programs on RandomCoffee, it already exists. Open your dashboard, filter on your newest hires, and you can see participation rates and their trend, the number of connections generated, and the people who have become less engaged.
Everything can be exported through result exports to sit next to your HRIS data. And if your new hires do not have a buddy program yet, our new hire onboarding page shows how to set one up in an afternoon, with newcomers added automatically from their start date.
Use day 60 to defend your onboarding budget
When budgets are reviewed, onboarding programs are often the first to be questioned, because their results are hard to show. A day-60 check-in gives you something concrete to bring to the table:
- the integration and clarity scores of each cohort, and where they improved;
- participation in buddy and cohort programs, with the number of connections created;
- early attrition compared with previous cohorts, as soon as the 90-day figures are in.
That is a stronger case than a welcome-day satisfaction score, and it ties your programs to the outcome leadership cares about: people who stay.
Key takeaway
Onboarding is hard to measure because the results that matter arrive late and the early signals flatter. Day 60 is the moment to look: the honeymoon is over, patterns are visible, and there is still time to act. Measure clarity, confidence and integration, watch participation as well as survey scores, and use what you find before day 90.
If your latest new hires have just passed their sixth week, now is the time to run the check-in. To see why connection carries so much weight in the results, read our article on belonging at work.
Frequently Asked Questions
When should HR measure onboarding success?
At several points, because each one answers a different question. Day 1 tells you how the welcome went. Day 60 is the most useful checkpoint: early impressions have faded, and there is still time to act before the 90-day mark. Retention at 12 months confirms the result, but it arrives too late to help the cohort you are measuring.
Why is new hire engagement so hard to measure?
Three reasons. The outcomes that matter, retention and performance, take months to show. Early satisfaction is inflated by the excitement of a new job, a pattern researchers call the honeymoon-hangover effect. And newcomers are often still on probation, so they rarely say openly that something is wrong. That is why behavioral signals, such as the meetings a newcomer actually attends, are worth tracking alongside surveys.
What should a day-60 onboarding survey ask?
Keep it to five questions on a 1 to 5 scale: whether the newcomer knows what is expected of them, whether they feel confident they can do the job, whether they know who to ask outside their own team, whether they have a colleague they can talk to openly, and whether they would recommend joining the company to a friend. The first three come from the three conditions research links to successful onboarding.
Should managers see individual answers to an onboarding survey?
No. Share results by cohort or by team, above a minimum group size, and keep individual answers with HR. Newcomers answer more honestly when they know their manager will not read their responses before a probation review. HR can still follow up individually with someone who reports low integration, with their agreement.
