The 18-Month Cliff: Why Good Employees Leave

18 month tenure cliff retention

Workzoom covers 18 month tenure cliff retention as part of the same platform that runs why good employees leave: on one employee record, with statutory rates maintained in the platform.

Most companies don't lose their best people to a better offer. They lose them to a month that should have had a conversation in it and didn't.

The resignation that blindsides you is rarely random. In hindsight, it's almost always predictable, and it's usually your strongest employees. The hire was good. The onboarding went fine. Eighteen months in, they're fully ramped, they know the clients, they're just hitting the point where they start paying back the investment. And then they leave.

Here's what most managers get wrong about the 18-month leaver. It's not a retention problem. It's a calendar problem. The decision gets made in a quiet stretch with no scheduled touchpoint, and by the time anyone notices, the book is already closed.

Good employees leave at 18 months because organisations invest heavily in the first 90 days and almost nothing in months 12 through 18. By that point the onboarding excitement has faded, but if a meaningful development conversation, a new challenge, or a visible path forward hasn't materialised, the next recruiter call becomes easy math. The 18-month departure isn't impulsive. It's a decision that was made quietly, three to six months earlier.

Why Good Employees Leave: The Arc Every Hire Follows

It's not random. There's a pattern, and once you see it, you can't unsee it.

Months 1–6. High engagement. Learning curve keeps things interesting. Every week brings something new. The hire is motivated partly by not wanting to look bad. Onboarding structures create natural touchpoints. Managers pay attention because the hire is still new.

Months 7–12. Settling in. The learning curve flattens. They're competent now. The natural check-ins from onboarding have dropped off. They start noticing things: how decisions get made, whether their ideas land, whether the promotion they were told was possible actually happens for people around them.

Months 13–18. The decision window. Excitement is gone. They know the job, they know the culture, they know the gaps. If something meaningful hasn't happened (a development conversation that went somewhere, a project that stretched them, a clear signal that progression is real) the calculus shifts. The question stops being "is this a good place to build a career" and starts being "is there somewhere better."

This is when the recruiter call gets returned.

Why the 18-Month Leaver Is Your Most Expensive

Not the longest-tenured employee. Not the most senior. The 18-month leaver.

Think about what you lose. Twelve-plus months of institutional knowledge: client relationships, internal context, process familiarity that took time to build. They're past the productivity dip that every new hire brings. They're not training anymore. They're contributing. And then they're gone, and you start over.

The other thing about 18-month leavers: they don't leave angry. The angry departure is actually easier to see coming and easier to learn from. The 18-month cliff is a quiet, rational decision. They liked it well enough. They just liked somewhere else slightly more. And "slightly more" is all it takes when the alternative is staying somewhere that never quite got around to investing in them.

Most organisations know, somewhere, which managers consistently lose people. They just don't want to formalise that knowledge because formalising it means acting on it. Crystal Murray, HR Generalist at Silvera for Seniors (400 staff, 30 buildings in Calgary), described the pre-Workzoom version of this honestly: "In the moments, no, we didn't want to know... out of sight out of mind situation for sure." The 18-month cliff is the same dynamic. You don't track tenure-based turnover patterns because tracking them means confronting them.

That's the 18-month cliff in institutional form. You don't track it because tracking it means confronting it.

The Structural Orphan Period

Most organisations have two structured retention moments: the 90-day review and the annual performance cycle. Everything between them is informal, manager-dependent, and wildly inconsistent.

That gap spans months three through twelve. Which means the entire lead-up to the decision window happens in an institutional vacuum. That stretch is when employees form their long-term read on whether this place is worth building a career in. No one is formally asking. No one is formally listening. The employee is drawing conclusions from ambient signals: whether their ideas get picked up, whether promotions happen for people like them, whether their manager seems to notice what they're doing.

Ambient signals are not a retention strategy. They're a lottery.

The problem isn't that managers don't care. It's that without structure, development conversations get displaced by everything urgent. Andy Woolley, Workzoom's CEO, frames it this way: the goal is "using technology to ensure that the key items are taking place with employees without it being a drain on the managers." Not to replace the conversation, but to make sure it actually happens.

Three Checkpoints That Close the Gap

You don't need a complex retention programme. You need three conversations at the right times, with the right questions.

The 6-Month Check

Not a performance review. A calibration. Six months in, the hire has enough context to give you an honest answer but hasn't made any decisions yet. That window closes.

The question that actually works: "What's something you expected to be doing by now that you're not?"

It's specific enough that they can't give you a non-answer, and it surfaces unmet expectations without putting them on the defensive. Most managers ask "how's it going" and get "good" back. This question doesn't let that happen. Silence at six months doesn't mean everything is fine. It usually means the employee has already adjusted their expectations downward and stopped mentioning it.

The 12-Month Check

The career conversation that needs to happen before the employee starts having it privately with themselves.

The question: "If a friend asked you what working here is actually like (not the interview answer) what would you say?"

This bypasses performance review mode. Asking someone to describe the experience to a hypothetical friend gets you the unfiltered version: the parts they've been editing out of every professional conversation. What you hear tells you whether they're still invested or whether they've mentally filed this job under "temporary."

The 18-Month Check

The one most organisations never have. The one that matters most.

The question: "Is there a version of this job that would be obviously better to you? Walk me through it."

Most managers avoid this because it feels like an invitation to complain. It isn't. It's the most useful information you can get. The employee who answers with something specific and achievable is telling you exactly what it takes to keep them. The employee who goes quiet or gives a vague answer has probably already moved on mentally, and now you know it early enough to do something.

What you're listening for is tense. Forward-looking answers ("I'd want to be leading this by then") versus past-tense answers ("I thought when I joined that..."). Past tense at 18 months is a signal, not a conversation starter.

The Quiet Leaver Is the Dangerous One

There are two types of voluntary departure. The angry resignation you can see coming: the blocked promotion, the bad review, the conflict that festered for months. And the quiet, rational exit where the employee simply ran the numbers and left.

The 18-month cliff is almost always the second kind. No drama. No triggering incident you can point to and fix. Just a gradual recalculation that landed on the wrong side of the ledger.

The angry leaver is easier to manage, paradoxically. There's a grievance. You can address it. You have something to work with. The quiet leaver has already done the work of separating their emotions from the decision. By the time they hand in notice, they're at peace with it. That's why counter-offers so rarely work: you're negotiating with someone who already closed the book. SHRM's workforce research consistently finds that the majority of employees who accept counter-offers leave within 12 months regardless. Not because they're being duplicitous. Because the offer addressed salary and the actual problem was something else entirely.

This is also why exit interview data fails specifically for the 18-month cliff. The quiet leaver doesn't have a grievance to articulate. They say "better opportunity" because it's true and because the real answer ("nothing bad happened, you just never gave me a reason to stay") is both harder to say and harder to hear.

You can't fix what you can't see. The three checkpoints exist to surface the quiet recalculation while it's still in progress, not after it's complete.

What This Requires Operationally

The three checkpoint conversations are simple in theory and consistently don't happen in practice. Not because managers are negligent. Because without a system that surfaces when a hire hits 6, 12, and 18 months and prompts the conversation, it gets displaced by whatever is urgent that week. In our 25 years of HR and payroll software, the pattern we see across client deployments is consistent: the companies that hold onto their best people aren't the ones with the richest perks. They're the ones whose system never lets a tenure milestone pass unnoticed.

This is the operational argument for HR infrastructure that tracks tenure milestones and flags them. Not sophisticated AI retention prediction. Just: this person hits 18 months on the 14th. Has the conversation happened? An HR system of record that owns hire dates, review history, and manager assignments can put that prompt in front of the right person a month early, every time.

You got into HR to build careers people don't want to leave, not to find out three months too late that one of them already had. When the 18-month conversation slips, it's not your memory that failed: your tools never surfaced the date. That visibility is what makes the playbook executable rather than aspirational.