The Language Barrier Hidden in Your Turnover Numbers

Nobody budgets for turnover the way they budget for labor.

Labor is a line item. You know what you're paying per hour, per shift, per head. Turnover is a cost that arrives sideways — recruiting fees, lost productivity, the two weeks before the new hire is useful, the institutional knowledge that walked out the door with the person who left. Most operations managers have a rough feel for the number. Very few have calculated it specifically, and almost none of them have broken it down by cause.

Knowing why people leave means knowing which departures were preventable

For facilities running multilingual workforces a measurable share of turnover links to communication gaps that nobody identified in time. The employee who left didn't file a complaint. She didn't explain it in the exit survey, which was in English, which she completed alone.

She found something else and moved on, and the record says "better opportunity," and the cost gets absorbed into the annual turnover rate like every other departure.

What the Research Says

Turnover cost estimates vary depending on who's doing the math, but the range that holds up across the research is consistent: replacing a frontline manufacturing or food processing employee costs somewhere between $3,000 and $7,000 when you account for recruiting, onboarding, and the productivity drag during ramp-up. At the higher end of that range — for skilled positions, for facilities with longer training periods, for operations where a new hire requires safety certification before they can work independently — the number climbs further.

The Society for Human Resource Management puts the average cost-per-hire across industries at over $4,700. The labor economics literature on frontline manufacturing turnover is consistent with that figure when you include indirect costs.

What the research doesn't say — because almost no one measures it — is how much of that cost is language-adjacent. That number requires looking at turnover data differently than most HR teams are set up to look at it.

Where Language-Adjacent Turnover Hides

Communication gaps in multilingual workforces don't generate incident reports or formal complaints. They generate confusion, and confusion generates departures. The mechanism is specific, and it plays out in the same three places across almost every facility that runs into this problem.

Onboarding

The first thirty days are where communication failures occur most. An employee who misunderstands her job expectations in week one is already at risk. If she can't ask a clarifying question in a language she's confident in — if the question requires navigating the complexity of asking a supervisor something in a second language she's still building — the question doesn't get asked. She proceeds on her best guess. Her best guess may be right. It may not be. Either way, she's starting from a position of uncertainty that a clear onboarding conversation would have resolved.

Studies on employee retention consistently show that the first 90 days are the highest-risk period for turnover. For multilingual employees who went through onboarding that didn't fully reach them, that window is even narrower.

Benefits enrollment

This one is underestimated. Benefits enrollment is a high-stakes, low-support moment for employees who aren't fully fluent in English. The forms are dense. The choices are consequential. The deadline is fixed. Many facilities run a single group enrollment session in English and consider the obligation discharged. The employee who couldn't fully follow the session, who couldn't confidently select the right health plan for her family, who ended up defaulting to whatever looked most familiar — she went home with unresolved questions about the most important financial decision her employer was asking her to make.

An employee who doesn't trust that she navigated enrollment correctly is an employee who isn't fully settled. She may find out months later that she made a coverage decision she didn't understand. She may never find out, but the uncertainty lingers. Either way, the relationship with her employer carries a gap that didn't have to be there.

Performance conversations

This is the highest-stakes communication failure of the three, and the one most likely to produce immediate turnover. A performance conversation that doesn't translate , where the employee hears something different than what HR intended to communicate, doesn't just create confusion. It creates a threat response. An employee who leaves a performance conversation unsure whether she's being warned, counseled, or terminated is not going to stay and ask for clarification. She's going to start looking.

This is the conversation most often handled through informal interpretation: the bilingual coworker pulled into HR's office, the supervisor who speaks enough of the employee's language to get through it fast, the translated version of a conversation that required impartiality and register accuracy to work correctly. When those elements are missing, the conversation lands differently than it was delivered. What HR meant as a documented coaching session becomes, in the employee's experience, something else entirely.

The Measurement Problem

Most facilities can't tell you what percentage of their turnover is language-adjacent, because they don't have the data infrastructure to find out. Exit surveys are in English. Exit interviews, when they happen at all, are conducted by HR staff in English. The feedback loop from multilingual employees runs through whatever informal translation infrastructure the facility happens to have — which is often the same bilingual coworker who was handling the performance conversation in the first place.

The measurement problem is solvable, but it requires changing a few things about how exit data is collected.

Multilingual exit surveys

If the exit survey is available only in English, you are measuring the English-speaking portion of your turnover. You are not measuring your multilingual workforce's departures in any way that captures their actual experience. Translating the exit survey — not with an app, with a qualified linguist who can maintain the meaning of the questions across languages — is the first step toward actual data.

Exit conversations in the right language

An exit interview conducted through an interpreter is more informative than an exit survey completed in a second language. Employees who left because of a communication gap are unlikely to identify that gap on a form they couldn't fully read. A conversation in their primary language, conducted by someone who is explicitly asking about it, will surface what the survey won't.

Disaggregating turnover by language

Most HRIS systems capture departure reason and departure date. Fewer capture primary language. If your facility isn't tracking primary language alongside turnover data, you can't run the analysis. Adding that field — and populating it — is the prerequisite for understanding whether language-adjacent departures are a material share of your total.

Calculating the Cost

Run a simple version of this for a 200-employee food processing facility with 20% annual turnover — a turnover rate that's on the low end for the industry.

40 departures per year at a conservative $4,000 replacement cost per head: $160,000 in annual turnover cost. If 20% of those departures are language-adjacent — a modest estimate for a facility running a predominantly Spanish or Haitian Creole workforce without formal language access infrastructure — that's 8 departures, or $32,000 in avoidable turnover cost per year.

A formal interpretation program covering onboarding, benefits enrollment, and performance conversations for a facility that size runs significantly less than that. The math isn't complicated. It's just math that most HR teams haven't been asked to run.

The 20% estimate is conservative. Facilities without structured language access in onboarding and benefits tend to see higher concentrations of early turnover among multilingual employees. If your first-90-day turnover rate among employees whose primary language isn't English is materially higher than your overall 90-day rate, the communication gap is a more likely explanation than coincidence.

What Changes When You Have the Infrastructure

Facilities that have built formal language access into their HR workflows see the difference in a few specific places.

Onboarding completion rates go up. When new employees can ask real questions in their first week — in a language where they have the vocabulary to ask them — they get answers that stick. The informal guesswork that drives early departures gets replaced with actual clarity about what the job is and what it requires.

Employees who make informed benefits decisions don't carry the quiet anxiety of having made choices they didn't understand. That's not a soft benefit. It's the difference between an employee who feels like her employer handled her like an adult and one who feels like she was handed a stack of forms and left to figure it out.

Performance conversations land as intended when a certified interpreter is in the room . The message HR delivers is the message the employee receives. That's the point. A warning is a warning. A coaching session is a coaching session. The ambiguity that drives reactive departures gets replaced with the clarity that keeps people in the room.

None of this is complicated. It's just a set of conversations that currently happen informally, through whoever is available, in whatever form felt like enough — and a set of conversations that can happen formally, through a process designed to make sure they work.

Understand Your Tunrover Data

If you have a multilingual workforce and your turnover data doesn't tell you why the multilingual employees are leaving, you don't have complete turnover data. You have partial data, and you're making staffing and retention decisions on a picture that's missing a significant portion of what's actually happening on the floor.

The question isn't whether language-adjacent turnover is costing your facility money. At any meaningful scale of multilingual employment, it is. The question is whether the number is big enough to show up in your budget conversation — and whether you have the data to make that case to whoever controls the line item.