Language Training for Employees: An Employer’s Guide

The first language training budget I ever helped approve was spent on the wrong thing. A company I worked with ran a small warehouse crew where most people spoke Spanish at home, and the safety briefing was being delivered in English to a row of polite, blank faces. English classes looked like the obvious answer.

They were not. Comprehension of a safety procedure is a legal obligation with a deadline of today, and language training is a project measured in years. The fix that week was a translated procedure and a bilingual supervisor walking the floor. The classes were still a good idea. They were just an answer to a different question, and buying them first would have left the actual risk sitting exactly where it was.

A year later the same budget line paid for something that worked beautifully: two quarters of weekly coaching for a support lead whose written English was costing us customers who misread her emails. One person, one task, one measurable result. This guide is about telling those two situations apart, and then buying well.

The guide covers what corporate language training is, the single condition that makes it worth funding, the safety and discrimination rules to check before you sign anything, the four formats and their published costs, how the providers differ, how to pay for it without a tax bill, and how to judge the result at ninety days. I build the people tooling for businesses without a dedicated HR person at FirstHR, which is the size of company this is written for.

What Corporate Language Training Is

Corporate language training is employer funded instruction that raises an employee's working proficiency in a specific language, scoped to the tasks the job requires rather than to general fluency. Providers sell the category under that name. Employers describe the same purchase as language training for employees, and both phrases return quotes from the same vendors.

The category runs in two directions, and mixing them up is the first way a budget goes astray. The inbound direction is English for employees whose first language is something else. The outbound direction is a second language for employees who sell to, support, or supervise people who speak it.

One more distinction decides which provider you should even call. Business English is workplace communication for someone who already has the language: emails, meetings, presentations, and difficult calls. General language acquisition is grammar and vocabulary from a low base. They are different products sold by different companies, and asking a business English provider to take someone from zero produces a polite quote and a poor result.

The scale of the inbound case surprises most employers. According to the United States Census Bureau, 77.7 percent of the population aged five and older spoke only English at home in the 2020 to 2024 American Community Survey five year estimates, which leaves 22.3 percent speaking something else.

In the workforce itself, the Bureau of Labor Statistics puts the foreign born share of the civilian labor force at 19.1 percent in 2025. This is not a topic that starts when you open an office abroad.

When Language Training Pays Off, and When It Does Not

Language training pays off when one repeated, identifiable job task is failing for language reasons and the person is otherwise good at the job. It fails, reliably and expensively, when it is bought as a general improvement for everybody.

That test sounds severe until you price the alternative. A translated procedure is paid for once. An interpreter for a quarterly all hands is paid by the meeting, not by the year. Language training is the most expensive instrument in the drawer per unit of change, so it should be reached for after the cheap instruments, not before them.

Row five deserves a moment because it arrives in almost every budget conversation. General improvement has no finish line, no baseline, and no owner, so it cannot be evaluated and therefore never gets cancelled. Ask which task is failing and the request either sharpens into something fundable or evaporates, and both outcomes are useful.

The other half of the honest answer is time. Language gain in adults is slow in a way that training budgets are not built to absorb, and the most common cause of disappointment is not a bad provider but a horizon set at one quarter.

Set those class hours against a realistic schedule and the planning becomes simple. One hour a week comes to roughly forty to fifty hours a year, which is a meaningful improvement in a narrow skill and nothing like a level change.

None of this argues against buying. It argues for buying against a task rather than against a level, and for a learning path that runs for years rather than a course that ends.

What the Law Says Before You Buy Anything

Two federal rules shape this purchase, and both of them bite before any training starts. The Occupational Safety and Health Administration (OSHA) requires safety training that employees can actually understand, and Title VII, a federal anti-discrimination law, limits what you can require of employees regarding language, accent, and fluency.

The safety rule is the one that turns a training decision into a compliance decision. In its training standards policy statement, OSHA takes the position that the words train and instruct mean presenting information in a manner employees are capable of understanding, so an employee who does not comprehend English must receive instruction in a language they do understand.

Where an employee's vocabulary is limited, the training has to account for that, and handing someone a manual to read does not satisfy the obligation if they cannot read it.

The practical consequence is the one I got wrong the first time. Language training does not meet a safety training duty, because the duty is owed now and the language gain arrives later. Translate, interpret, and demonstrate today, then run the classes on their own timeline. Everything else in your OSHA obligations and in safety training generally assumes comprehension has already been solved.

The discrimination rules are less familiar and easier to breach by accident. The Equal Employment Opportunity Commission treats an English fluency requirement as permissible only where it is required for effective performance of the position, and expects the assessment to be made role by role rather than applied across dissimilar jobs.

An employment decision based on accent stands only where effective spoken English is required and the accent materially interferes with performance, supported by evidence rather than assertion.

Three practical habits follow. First, write the language requirement into the job description as a task the role performs, such as reading and following written work instructions, rather than as a blanket demand for fluency. Second, apply it only to the roles that genuinely perform that task.

Third, keep language training voluntary wherever it is not tied to a specific job duty, because Title VII exposure grows the moment training starts to look like a condition of employment aimed at one national origin group.

Four Ways to Buy It

Four formats cover the market, and they differ far more in what they can change than in what they cover. Price per hour is a poor guide on its own, because the cheapest hour is the one that moves nothing.

Headcount decides the crossover between formats. Coaching costs five times as much for five learners as for one, while an app seat costs less per month than a single coaching hour, so adding learners strains the budget far less. Group classes sit between the two and are the cheapest per contact hour of the four, provided you can actually assemble a group at one level, which is the constraint most small employers underestimate.

There is a fifth option that almost nobody prices, because it is publicly funded rather than sold. State administered adult education programs run on formula grants under Title II of the Workforce Innovation and Opportunity Act (WIOA), and English language acquisition is one of the activities those grants pay for.

The Act also defines workplace adult education and literacy activities, meaning classes an eligible provider runs in partnership with an employer, at the workplace or off site. The Department of Education tells individuals and local providers to check with their state office for information on how adult education and literacy programs are funded.

For a frontline employee starting from a low base, a nearby publicly funded class is worth a phone call before you price anything commercial, since the terms are set by the state provider and your own cost may come down to paid release time.

The Provider Landscape

The providers in this market look similar on their home pages and sell genuinely different things underneath. The useful question is not which is best but which category of product you are being quoted for.

Almost every enterprise provider on that list quotes rather than publishes, which makes comparison harder than it should be. Four questions cut through it.

Ask each provider what the price per learner per year is, not per license band. Ask who does the placement testing and whether it is free. Ask what the trainer replacement policy is if the learner and the trainer do not get along. And ask what exactly appears in the report you get at the end, since a completion percentage and a level change are very different pieces of evidence.

Ask for a pilot as well. A provider confident in the product will run one learner for one term at list price, and a provider that will only sell an annual block for the whole team is telling you something about its renewal rates. The same discipline applies to any e-learning platform purchase, and for the same reason.

What Corporate Language Training Costs

Published rates start at roughly $8 per learner per month for a self paced license and run to about $45 an hour for in person one to one instruction. The table below uses only the lowest figures the providers themselves publish, so read them as the floor of the market rather than as a quote.

Two costs never appear on a quote and often exceed it. The first is paid class time, covered in the next section, which for a weekly session across a year is a meaningful share of one person's salary. The second is coverage: someone runs the counter, answers the phone, or watches the line while the learner is in class, and on a small team that person is usually a manager.

Against those, put the cost of the failure you are trying to fix. If a misread instruction costs an hour of rework twice a week, that is roughly a hundred hours a year, which prices most of this table without argument. This is the same arithmetic that governs any training program you start from scratch, and language training is simply the version where the bill is larger and the timeline is longer.

Paying for It Without Creating a Tax Bill

A written educational assistance plan lets you pay for language training tax free up to an annual cap. Paying for it informally does not, unless the training is job related enough to count as a working condition benefit. This is the cheapest thing in the whole exercise to get right, and the most commonly skipped.

Under section 127 of the tax code, an employer can exclude up to $5,250 per employee for calendar year 2026 from the employee's gross income for educational assistance, and the cap is indexed for inflation from 2027.

The Internal Revenue Service states plainly that the payments do not have to be for work related courses, and the only content exclusion is for sports, games, and hobbies unless the course relates to your business or forms part of a degree program. A language course clears that bar comfortably in either direction.

The working condition route is worth knowing. Where the training maintains or improves the skills a job already requires, Publication 15-B treats it as excludable, on the basis that the cost would have been deductible as a business expense had the employee paid it.

Job related training can therefore stay out of wages even when the assistance goes above the section 127 cap. If you already run tuition reimbursement, language courses usually slot into the existing plan without a new document. Nothing stops an employer from using both routes in the same year.

The other money question is whether the class counts as paid time, and it catches employers who assume a class is not work. Federal wage and hour rules treat training time as hours worked unless attendance is outside regular hours, voluntary, unrelated to the job, and free of productive work.

The narrow exception in the federal regulations at 29 CFR 785.31 covers an employer established program of instruction corresponding to courses offered by independent schools or colleges, where voluntary attendance outside working hours is not hours worked even if the course is job related and the employer pays for it.

My advice is to stop trying to fit through that exception. Schedule the class inside the working week, pay for the time, and accept the cost, because attendance is the single largest determinant of whether the money produces anything. A program that is free of payroll cost and half attended is more expensive than a paid one that runs to completion.

Running the Rollout

A language program lives or dies on three decisions made before the first class: what task it is aimed at, what level the learner starts from, and where the session sits in the week. Everything else is administration.

Those seven steps fit on one page, which is what the worksheet below is. Fill it in before you ask any provider for a quote, and the conversation with the sales team becomes much shorter.

The worksheet forces the two things that get skipped: a counted baseline for the failing task, and the learner's own account of what is hard. Block eight is the ninety day review, and it is deliberately on the same page as block one so that the comparison is unavoidable.

For anyone joining after the program is running, put the arrangement into onboarding rather than leaving it to be discovered. A new hire who learns in week one that classes exist, are paid, and are offered on a task basis is far more likely to ask than one who hears about it in month seven.

How to Tell Whether It Worked

Judge the first ninety days on the task that started the program and leave level movement as a twelve month question. Level is the provider's metric and it moves slowly. The task is yours and it moves first.

Take the baseline before the first class even if it is rough, because a remembered baseline drifts toward whatever makes the decision look good. Two of them, the task count and attendance, need nothing more than a weekly tally, and the rest are already visible in how the work runs.

Resist the temptation to score satisfaction. Learners enjoy good teachers and report high satisfaction long before anything changes at work, which is the classic reaction level result that tells you nothing about behavior. Compare against your own previous number rather than any published benchmark, since the population in those benchmarks looks nothing like your team.

Common Mistakes

The failures repeat, and most of them come from buying the format the category defaults to rather than the one the situation calls for.

The eighth row is the one that quietly does the most damage. When only the employees whose first language is not English receive training, the message is that they are the deviation and everyone else is the standard. Offering the supervisor coaching in the other direction, or simply some vocabulary for the trade, costs little and changes how the whole program is received.

If you do one thing this week, write the failing task in one sentence and count how often it happens over the next four weeks. A task you can name and count is worth a coaching budget. A task you cannot is a reason to reach for the cheap instruments first, or to budget the purchase openly as a benefit.

Frequently Asked Questions