HomeThe ProblemROICase StudiesAboutResourcesPricingContact Start Hiring →
Guide · Screening

How to Test a Remote Hire’s English Before You Hire

A practical method for checking the English of overseas candidates: which CEFR level each role needs, what country averages do and don’t tell you, a four-step test, and a scoring rubric you can copy.

The CEFR levels, with the matching EF EPI scores

  1. C2Proficient userUnderstands virtually everything heard or read; expresses themselves very fluently and precisely.EF EPI 700–800 · Very high
  2. C1Proficient userExpresses themselves fluently and spontaneously; uses language flexibly for professional purposes.EF EPI 600–699 · Very high
  3. B2Independent userUnderstands technical discussions in their field; interacts with native speakers without strain.EF EPI 500–599 · High (550–599) / Moderate (500–549)
  4. B1Independent userHandles familiar matters at work; describes experiences and gives brief reasons.EF EPI 400–499 · Low (450–499) / Very low
  5. A2Basic userCommunicates in simple, routine tasks on familiar topics.EF EPI 300–399 · Very low
  6. A1Basic userUses familiar everyday expressions and very basic phrases.EF EPI 200–299 · Very low

How do you test a remote candidate’s English? Decide the level the role needs on the CEFR scale (usually B2 or C1 for client-facing work), then test in three layers: a standardized CEFR-aligned test for a baseline, a written task taken from the real job, and a live spoken interview or role play. Score every candidate on the same rubric, and weight the skills the job actually uses. Country rankings help you choose where to look, not whom to hire.

Step 1: Set the level by role, not one bar for everyone

The CEFR, published by the Council of Europe, describes six levels from A1 to C2. Most client-facing remote roles sit at B2 or C1, but which skill matters most changes by job: a developer reads and writes all day, while an appointment setter lives on the phone. Pick a role for a starting point:

Suggested starting level

B2+ spoken, ideally C1

What to test

Live call role play with objections; a short follow-up email

How we hire for this role →

See all roles as a table
RoleSuggested starting level (CEFR)What to test
SDR or appointment setterB2+ spoken, ideally C1Live call role play with objections; a short follow-up email
Customer supportB2+ spoken and writtenMock ticket reply plus a live de-escalation call
Executive assistantC1 written, B2+ spokenDraft an email on your behalf; summarise a messy thread
DeveloperB2 reading and writing; B2 spoken for stand-upsExplain a technical decision to a non-technical stakeholder; write a pull-request description
Virtual assistantB2 written and spokenReal admin task with written instructions; a short call to confirm details
BookkeeperB2 reading and writtenWrite a client query about a missing receipt; read and act on a written policy

Suggested starting points for remote roles working with English-speaking teams and clients. Raise the bar for senior and client-facing seats; lower it when work is mostly written to templates.

Step 2: Use country averages to choose where to look

The EF English Proficiency Index 2025 is based on 2.2 million test-takers of the EF SET, an adaptive test of reading, listening, speaking and writing. It’s useful for choosing a market, as long as you remember that a national score is an average. The skill breakdown is more useful still: in 11 of these 12 markets reading is the strongest skill, and in 10 the weakest is writing or speaking, the two a remote hire uses most.

11 of 12markets where reading is the strongest English skill
10 of 12where the weakest skill is writing or speaking
112 ptswidest gap between best and weakest skill (Bulgaria)
EF EPI 2025 scores by skill: reading, listening, writing and speaking, by countryRomania: reading 621, listening 601, writing 596, speaking 534; Poland: reading 612, listening 597, writing 557, speaking 518; Bulgaria: reading 613, listening 594, writing 534, speaking 501; Hungary: reading 603, listening 587, writing 589, speaking 523; Czech Republic: reading 598, listening 579, writing 537, speaking 498; Argentina: reading 592, listening 565, writing 526, speaking 489; Philippines: reading 573, listening 567, writing 603, speaking 539; Morocco: reading 532, listening 460, writing 491, speaking 491; Brazil: reading 516, listening 468, writing 442, speaking 464; Colombia: reading 509, listening 457, writing 465, speaking 470; Egypt: reading 469, listening 451, writing 403, speaking 453; Mexico: reading 455, listening 422, writing 399, speaking 413ReadingListeningWritingSpeaking400440480520560600640RomaniaPolandBulgariaHungaryCzech RepublicArgentinaPhilippinesMoroccoBrazilColombiaEgyptMexico
EF English Proficiency Index 2025 country fact sheets: scores by skill. The grey bar runs from each country’s weakest to its strongest skill. Countries are those where Talentige recruits.
See the data as a table
CountryOverall (rank of 123)ReadingListeningWritingSpeakingGap, best vs weakest skill
Romania605 (#11)62160159653487
Poland600 (#15)61259755751894
Bulgaria594 (#18)613594534501112
Hungary590 (#22)60358758952380
Czech Republic582 (#23)598579537498100
Argentina575 (#26)592565526489103
Philippines569 (#28)57356760353964
Morocco492 (#68)53246049149172
Brazil482 (#75)51646844246474
Colombia480 (#76)50945746547052
Egypt458 (#89)46945140345366
Mexico440 (#103)45542239941356

So a reading-heavy test can flatter a candidate whose spoken English isn’t ready for calls. Test what the job uses, live. For time-zone and cost trade-offs between these markets, see nearshore vs offshore.

Step 3: Test in four layers

  1. 1
    A standardized, CEFR-aligned testA quick, objective baseline across reading and listening, ideally speaking and writing too. EF says its free EF SET is designed to classify test-takers into the six CEFR levels. Use it to filter, not to decide.
  2. 2
    A written task from the real jobA client email, a ticket reply, a pull-request description or a summary of a messy thread. Give everyone the same prompt and time limit.
  3. 3
    A live spoken interviewUnscripted follow-up questions show comprehension and fluency far better than prepared answers. Ask them to explain something from their own work.
  4. 4
    A role-specific simulationA sales call with objections, a de-escalation call, or explaining a technical decision to a non-technical stakeholder. This is where readiness shows.
Abstract illustration of speech bubbles rising up a six-step ladder, with a sound wave and a written line beside them

Step 4: Score everyone on the same rubric

Score each criterion from 1 to 5 during the interview, not afterwards from memory. Two interviewers scoring separately and then comparing catches most bias.

Criterion1 · Not ready3 · Workable5 · Strong
ComprehensionAsks you to repeat often; misses the point of questionsFollows normal-speed conversation; asks to clarify only on detailFollows fast, idiomatic speech and implicit meaning
FluencyLong pauses; rehearsed answers break downSpeaks at length with some searching for wordsSpeaks spontaneously, with no obvious searching
AccuracyErrors change the meaningErrors are noticeable but don’t block meaningRare errors, self-corrected
RangeSimple vocabulary only; avoids the topicExplains and gives reasons on work topicsPrecise, nuanced; adapts tone to the listener
WritingMessages need rewriting before they go outClear, mostly correct emails on familiar topicsClient-ready writing, including on complex topics

Who can validate English for you

Test providers

Standardized tests that report a CEFR level. Fast and comparable across candidates, but they test general English, not your job.

Recruitment partners

A recruiter that tests written and spoken English as part of screening, so you only meet candidates who passed. Talentige runs written and oral English tests for every candidate, plus cold-call role plays for sales roles.

Your own team

Your engineers, sales leads or managers judging the role-specific simulation. Nobody knows better what “good enough” sounds like on your team.

Red flags and fair testing

Answers that sound memorised

Fluent introductions that fall apart on the first follow-up question.

Camera off, or a long delay before every answer

Ask for a second short call; it may be the connection, or it may not.

Written work that doesn’t match the interview

Run a short written task live, in the call, as a check.

Judging the accent instead of the English

Score comprehension, fluency, accuracy and range. An accent alone doesn’t predict how someone works.

Prefer to meet only candidates who’ve passed?

Talentige tests every candidate’s written and spoken English, plus role skills, before you meet anyone. You interview 3–5 finalists.

Book a free hiring call →

Pay nothing until you hire · 90-day replacement guarantee

Hiring in a specific market? See Latin America, the Philippines, Morocco, Egypt or Eastern Europe, or start with how to hire international employees.

Questions about testing English

How do I test the English of a remote hire?
Combine three things: a standardized, CEFR-aligned test for a baseline, a written work sample from the actual job, and a live spoken interview or role play. Score each against the same rubric for every candidate, and set the pass level by role, not one bar for everyone.
What level of English do I need for a remote role?
It depends on the job. Customer-facing calling roles usually need strong spoken English, around B2 and above on the CEFR, while roles that write on your behalf need strong written English. Reading-heavy technical roles can work with B2 reading and writing. Use the role picker above as a starting point and adjust it to your team.
What is the CEFR?
The Common European Framework of Reference for Languages, published by the Council of Europe. It describes language ability on six levels, from A1 (basic) to C2 (proficient), and most English tests report results against it.
Are country English rankings useful for hiring?
As a starting point only. The EF English Proficiency Index compares national averages; EF itself notes that a proficiency band only reflects the average test-taker. Within any country you’ll find candidates far above and below the average, so test every candidate.
Which skill do candidates usually struggle with most?
In the EF EPI 2025 fact sheets for the 12 markets in this guide, reading was the strongest skill in 11 of 12 countries, and the weakest skill was writing or speaking in 10. Test the skills the job uses live, especially speaking and writing, rather than relying on a reading-heavy test.
What kinds of partners can validate English fluency for technical hires in Latin America?
Three kinds: standardized test providers that report CEFR levels, recruitment partners that test spoken and written English as part of screening, and your own engineers, who can judge technical communication in a live interview. Talentige runs written and oral English tests on every candidate before you meet them, alongside technical tests for developers.
Should I reject candidates with a strong accent?
Judge whether you can understand them easily and whether they understand you, not the accent itself. Comprehension, fluency, accuracy and range predict how well someone works with your team; an accent on its own doesn’t.

Sources

  1. Council of Europe, Common European Framework of Reference for Languages, Global scale – Table 1 (CEFR 3.3): Common Reference levels. coe.int
  2. EF Education First, EF English Proficiency Index 2025 report: based on data from 2.2 million EF SET test-takers in 2024; the EF SET is an online, adaptive test of reading, listening, speaking and writing designed to classify test-takers into the six CEFR levels and is free; EF EPI score bands and CEFR correspondence; a proficiency band indicates the skills of an average test-taker. ef.com
  3. EF Education First, EF EPI 2025 country fact sheets (overall score, global rank and scores by skill) for Romania, Poland, Bulgaria, Hungary, Czech Republic, Argentina, Philippines, Morocco, Brazil, Colombia, Egypt and Mexico. ef.com
  4. Talentige screening process (as of October 2026): written and oral English tests, communication tests and role-specific tests for every candidate.