How to Build Multilingual Typing Tests That Actually Work

A customer service rep in Miami fields calls in English and Spanish. A legal transcriptionist in Montreal switches between French and English briefs. A virtual assistant in Manila toggles between Tagalog emails and English Slack messages. These aren't edge cases anymore. They're the reality of a workforce where bilingual and multilingual skills are a baseline expectation, not a bonus.

But here's the problem most recruiters don't think about until it's too late: a standard English typing test tells you almost nothing about how fast or accurately a candidate types in their second (or third) language. Different languages have different word lengths, character frequencies, diacritical marks, and keyboard layouts. Testing a Spanish-English bilingual candidate with an English-only passage is like evaluating a swimmer by watching them run.

According to research from the U.S. Department of Education, over 68 million people in the U.S. workforce speak a language other than English, with bilingual demand concentrated in customer service, sales, healthcare, and administrative roles. If you're hiring for any of these positions, you need a typing assessment strategy that reflects the actual work your candidates will do.

This guide walks you through setting fair WPM benchmarks across languages, navigating keyboard layout differences, and keeping your multilingual assessments legally defensible. And if you want to start building custom bilingual typing tests right now, you can sign up for TypeFlow and create your first assessment in minutes.

Setting Fair WPM Benchmarks Across Different Languages

One of the biggest mistakes recruiters make with multilingual typing assessments is applying a single WPM benchmark across all languages. A 60 WPM standard might be perfectly reasonable for English, but that same number can be wildly unfair or meaninglessly easy depending on the target language. Understanding why requires a quick look at how languages differ at the keystroke level.

Why WPM Doesn't Translate Directly Between Languages

Words per minute is calculated by dividing total characters typed by five (the standard "word" length) and then dividing by the elapsed time in minutes. The problem? Average word length varies dramatically between languages. English averages about 4.7 characters per word. German averages closer to 6.3 characters because of compound nouns like "Krankenversicherung" (health insurance). Spanish falls around 5.2 characters per word, while French sits near 5.1.

This means that a typist producing identical keystrokes per minute will register a lower WPM in German than in English, simply because each "word" requires more keystrokes. A candidate who types 55 WPM in German might actually be producing the same raw output as someone typing 65 WPM in English.

Then there are diacritical marks. French requires accents (é, è, ê, ë), Spanish uses tildes (ñ) and inverted punctuation (¿, ¡), and Portuguese adds cedillas (ç). Each of these characters typically requires an extra keystroke or a modifier key combination, which adds friction and slows raw WPM without reflecting any deficit in actual typing ability.

Practical Benchmark Adjustments by Language

Here's a framework you can use to set fair, role-appropriate benchmarks for your most common bilingual hiring scenarios:

These ranges account for the character-per-word differential and the additional keystroke burden of diacritical marks. The key principle: drop your baseline by roughly 5-10 WPM for languages with longer average words or frequent special characters.

For accuracy benchmarks, the adjustment is smaller. Most recruiters should hold a consistent accuracy standard of 95-97% across languages, since accuracy reflects attention to detail regardless of the language being typed. However, if your test passages include heavy use of accented characters, consider allowing a 1-2% grace margin for candidates who are typing on unfamiliar hardware.

When you build tests in TypeFlow, you can configure separate pass criteria for each test you create. That means you can set a 55 WPM / 96% accuracy threshold for your Spanish customer service test and a 60 WPM / 97% threshold for the English version of the same role. Browse industry-specific test templates to get a head start on passages tailored to medical, legal, customer service, and data entry contexts, then customize the pass criteria for each language variant.

Keyboard Layouts, Test Passages, and the Details That Trip Up Recruiters

Benchmarks are only half the equation. The other half is making sure your test environment doesn't accidentally penalize candidates for factors that have nothing to do with their actual typing ability. Keyboard layouts and passage selection are the two areas where things go wrong most often.

QWERTY, AZERTY, QWERTZ, and Beyond

Most recruiters assume everyone types on a QWERTY keyboard. That assumption breaks down fast once you're hiring across borders. French typists commonly use AZERTY, where the positions of A/Q and Z/W are swapped, and numbers require a Shift key. German typists use QWERTZ, where Y and Z trade places. Latin American Spanish keyboards add dedicated keys for ñ, ¿, and ¡ but rearrange some punctuation.

Why does this matter for remote assessments? Because candidates taking your test from home will be using their native keyboard layout. If your test passage is loaded with characters that are easy to reach on a QWERTY layout but awkward on AZERTY, you're measuring keyboard ergonomics, not typing proficiency.

Here's what you should do:

• Never require candidates to switch keyboard layouts. Your test instructions should explicitly tell candidates to use whatever layout they normally work with.

• Match your test passage language to the expected keyboard layout. A French passage should be tested with the understanding that the candidate is on an AZERTY keyboard.

• Avoid cross-language symbol-heavy passages. If your English test has lots of @ and # symbols, don't reuse that passage structure for a French test where those symbols live in different positions.

• Account for layout switching time in bilingual tests. If you're testing a candidate in both English and Spanish in the same session, build in a brief buffer between passages so they can mentally (and physically) adjust.

Choosing Test Passages That Reflect Real Work

Generic "the quick brown fox" passages are useless for multilingual hiring. Your test passages need to mirror the actual content your employees will type on the job. This is where specificity pays off.

For a bilingual customer service role at an insurance company, your English passage might read:

"Thank you for calling Pacific Health Insurance. I can see your claim was submitted on the 14th. Let me check the status for you. It looks like we need an updated referral from your primary care physician before we can process the reimbursement."

And the Spanish passage for the same role:

"Gracias por llamar a Pacific Health Insurance. Puedo ver que su reclamo fue enviado el día 14. Permítame verificar el estado. Parece que necesitamos una referencia actualizada de su médico de cabecera antes de poder procesar el reembolso."

Notice how the passages test the same scenario, the same vocabulary domain, and similar sentence complexity. This parallelism is what makes your results comparable across languages.

Poor passage selection introduces noise. If your English passage is a conversational email and your Spanish passage is dense legal text, you're not comparing apples to apples. You're comparing apples to courtroom transcripts.

TypeFlow lets you create custom tests with your own passages, set configurable durations, and define separate pass criteria for each version. You can build one test for English customer service, another for Spanish customer service, and then send both to the same candidate via unique shareable links. After they complete the assessments, the TypeFlow dashboard gives you side-by-side performance metrics, including WPM, accuracy, keystroke analysis, and any security violations like tab switching or paste attempts.

Keeping Multilingual Typing Assessments Legally Defensible

Here's where the stakes get serious. A typing test that inadvertently discriminates against candidates based on their native language or national origin isn't just bad practice. It's a legal liability. Employment assessments in the U.S. must comply with Title VII of the Civil Rights Act and the EEOC's Uniform Guidelines on Employee Selection Procedures. Similar frameworks exist in the EU, Canada, and other jurisdictions.

The core legal principle is straightforward: your assessment must be job-related and consistent with business necessity. A typing test is defensible when it measures skills the employee will actually use on the job. It becomes legally vulnerable when it measures something irrelevant, or when its design systematically disadvantages a protected group.

Three Principles for Defensible Multilingual Assessments

1. Test what the job actually requires.

If the role requires typing in Spanish 70% of the time and English 30% of the time, weight your assessment accordingly. Don't give equal weight to both languages if the job duties aren't equal. Document the job analysis that led to your testing criteria. This documentation is your first line of defense if a candidate ever challenges the assessment.

2. Use equivalent, not identical, test content.

As discussed above, the same WPM benchmark across different languages isn't "equal." It's discriminatory by design because it ignores the structural differences between languages. True equivalence means adjusting for word length, character complexity, and keyboard layout differences so that candidates of equal ability produce comparable results regardless of the language being tested.

For a deeper dive into building bias-resistant assessments from the ground up, check out this guide on building legally defensible, bias-resistant typing tests.

3. Apply consistent proctoring standards.

If you're monitoring for cheating in your English test, apply the same monitoring to your Spanish test. Inconsistent proctoring creates the appearance of differential treatment. TypeFlow's security monitoring works identically across all tests, detecting tab switches, paste attempts, focus loss, and suspicious typing patterns regardless of the passage language. This consistency is exactly what auditors and legal teams look for.

Documentation Checklist for Legal Compliance

Keep a record of the following for every multilingual assessment you administer:

• Job analysis showing which languages are required and in what proportion

• Rationale for WPM and accuracy benchmarks in each language

• Source and rationale for test passage selection

• Instructions provided to candidates about keyboard layout and test environment

• Accommodation procedures for candidates who request them

• Aggregate pass/fail rates broken down by language to monitor for adverse impact

That last point deserves emphasis. If your Spanish typing test has a 30% pass rate while your English test has a 70% pass rate, you have a potential adverse impact problem. The four-fifths rule (used by the EEOC as a rule of thumb) says that the selection rate for any group should be at least 80% of the rate for the highest-performing group. Monitor your results and adjust benchmarks if the data shows a disparity that can't be explained by legitimate job requirements.

Putting It All Together: A Step-by-Step Implementation Plan

You've got the benchmarks, the passage strategy, and the legal framework. Now here's how to pull it all into a working system that scales with your hiring.

Step 1: Audit your current roles. Identify every position that requires typing in more than one language. Document the language split (e.g., 60% English, 40% French) and the type of content being typed (emails, transcription, data entry, chat support).

Step 2: Build language-specific test passages. Write or source passages that mirror real job tasks in each required language. Aim for 200-300 words per passage, with vocabulary and complexity that match the role. Use your own company communications as source material when possible.

Step 3: Set adjusted benchmarks. Use the framework from this article to set WPM targets that account for language-specific differences. Start with the English benchmark your team already uses, then adjust downward by 5-10 WPM for languages with longer words or more diacritical marks.

Step 4: Create and distribute tests. Build separate tests for each language in TypeFlow, configure your pass criteria, and send unique test links to candidates. For high-volume hiring, TypeFlow's Professional and Enterprise plans support bulk email invitations with CSV upload, so you can send bilingual assessments to hundreds of candidates at once.

Step 5: Analyze results comparatively. Don't just look at raw WPM. Compare each candidate's performance against the language-specific benchmark. A candidate who scores 52 WPM in Spanish (against a 50 WPM target) and 58 WPM in English (against a 60 WPM target) is strong in Spanish but borderline in English. That matters for role placement.

Step 6: Monitor and iterate. Track pass rates by language over time. If you see persistent disparities, revisit your benchmarks and passage difficulty. Continuous monitoring is both a best practice and a legal safeguard.

Multilingual typing assessment doesn't have to be complicated. It does have to be intentional. The recruiters who get this right will hire faster, hire better, and build teams that actually reflect the multilingual reality of their customer base.

Ready to build your first bilingual typing test? Sign up for TypeFlow and start creating custom, proctored assessments with adjustable benchmarks, security monitoring, and detailed analytics for every language your team needs.