How to Assess Employee English Levels, Practically
By Elena Vasquez, Business English Trainer · Jun 4, 2026 · 9 min read
What is the most practical way to assess employee English levels?
Put everyone on the CEFR scale, measure spoken English rather than self-reports, and pair one placement measurement with continuous assessment from real practice.
Written tests and dusty certificates tell you what somebody once knew. A spoken assessment tells you what they can do in a meeting on Tuesday.
I have spent fifteen years doing these assessments in meeting rooms from Munich to Madrid. HR usually inherits a spreadsheet column called English holding words like good, fluent and native-ish.
None of those are data. The fix costs less than most teams assume, and it starts with refusing to accept self-description as measurement.
• Adopt CEFR as the single shared scale
• Measure speaking, not reading or grammar drills
• Verify every inherited level with your own ears
• Re-measure quarterly, and act on what moves
Why do self-reported levels and dusty certificates mislead?
Both describe a best day rather than an average one. People rate their English against the finest conversation they ever had, not the one where a supplier is angry and talking fast.
An operations director in Rotterdam once handed me a list of staff marked B2 or better. After short speaking interviews, a good number were nowhere near it. Nobody had lied, exactly.
Certificates have the opposite problem: they were probably accurate once. Language decays without use, and in my experience it decays fastest in the skill business needs most, spontaneous speaking.
There is a quiet incentive problem too. Nobody downgrades themselves during hiring, and the self-assessed line then follows the person into your HR system and hardens into fact.
• Self-ratings measure confidence more than capability
• Old test formats weight reading and listening over speech
• A six-year-old score describes a six-year-old speaker
• Treat every unverified level as a hypothesis
Why should CEFR be your shared scale?
Because it is the one scale every serious test, vendor and platform already maps to. CEFR runs from A1 to C2 and describes what a person can do, not what they scored.
That lets you finally compare a certificate, a placement interview and an app score on a single axis. Our article on CEFR levels walks through each band with workplace examples.
Do not let the scale become a fetish. The moment level labels drive pay or status, people start gaming the assessment instead of using the language.
• B1 handles routine work with visible effort
• B2 works independently in meetings, the usual minimum
• C1 negotiates and persuades under pressure
• C2 is rarely a business requirement worth funding
What do standardised tests actually buy you?
They buy you proof, not operating data. A recognised test is externally validated, defensible in an audit and comparable across candidates and countries.
When a regulator, a client contract or a relocation case needs a certified level, nothing else substitutes for it. Let me be fair to the instrument on that point.
For internal capability planning, though, it is an expensive snapshot. You pay per head, schedule people away from their work, and learn what each person managed on one morning.
Six months later the number is stale and the money to refresh it has gone. I have watched companies run a testing round, file the results and never act on them.
Are instructor placement interviews enough?
They are a strong instrument with one blind spot, which is timing.
A trained assessor listening to live speech and mapping it to CEFR is exactly the measurement I have argued for here. Most corporate language training providers build their courses on such an intake assessment.
The limitation is that it happens once, at enrolment. It tells you where somebody started, and nothing about whether anything has changed since.
So if your question is which teams are progressing and which have stalled, an intake interview cannot answer it, however skilled the assessor. Plan the re-measurement question separately.
What does continuous assessment from real speaking practice look like?
It means measuring people while they practise, instead of interrupting work to test them. This barely existed when I started training.
Lucida for Business is how we do it. Employees hold spoken conversations with an AI coach, Lucy, in realistic scenarios such as meetings, presentations and interviews.
Scores are CEFR-aligned and tracked over months, so assessment stops being an event and becomes a byproduct of practice, with no test-day nerves and no scheduling round.
A caveat, because I promised honesty: an app measures only what happens inside the app. If an employee never opens it, you have no data on them, which is a signal but a different one.
Adoption is a management question, not a measurement question. No vendor, Lucida included, removes the need for a manager who cares whether the team actually practises.
• Feedback on pronunciation, grammar, vocabulary, fluency, pace and filler words
• CEFR-aligned speaking scores that move over time
• A dashboard showing the CEFR spread across a team
• Per-employee assessments without a test day to organise
What should you do with the results?
Act selectively. The point of assessment is triage, not a training round for everybody, which flatters nobody and drains the budget.
A B1 engineer documenting systems in their own language may be fine for years. A B1 account manager defending a renewal with a London client is a revenue risk this quarter.
Never use assessment results punitively. The first time a level score turns up in a performance review, honest engagement with the whole thing dies, and it does not come back.
• Map each role to the CEFR level the work genuinely demands
• Compare measured level against the role requirement, per person
• Fund training where the gap touches revenue, safety or clients
• Leave adequate speakers alone; goodwill matters and budgets are finite
How often should you re-measure?
Quarterly, and not more often. Meaningful movement between CEFR levels takes months of consistent practice, so weekly numbers are mostly noise wearing a suit.
A quarterly rhythm matches how most teams already plan budget and headcount. It is frequent enough to catch a stalled cohort before a year of training spend quietly evaporates.
This is where continuous measurement pays for itself. The quarterly review becomes a report you open rather than a testing round you organise.
Compare the CEFR distribution against last quarter, check the people whose scores have gone flat, reallocate. That fits in one meeting, which is as much ceremony as this deserves.
Frequently asked questions
What is the most reliable way to test an employee’s English level?
A speaking-based assessment mapped to the CEFR scale. Speaking fails first in meetings and decays fastest, yet most quick tests measure reading and grammar. Use a live placement interview or a tool that scores actual spoken performance, and treat self-reported levels as hypotheses until verified.
Are old English certificates a valid indicator of current level?
Treat them cautiously. A certificate describes performance on the test date, in that test’s format. Ability declines without regular use, and speaking declines faster than reading. If the certificate is several years old and the employee has not used English since, assume the speaking level is lower.
Do corporate language training providers include level assessments?
Most include an instructor-led placement assessment at enrolment. A trained human evaluating live speech is a strong instrument, though details vary by contract, so confirm specifics with any vendor you shortlist. The main limitation is that placement happens once, so it cannot show you progress afterwards.
Can an app really assess CEFR speaking level?
Yes, within honest limits. Lucida has employees hold spoken conversations with an AI coach and produces CEFR-aligned speaking scores tracked over time, with feedback on pronunciation, grammar, vocabulary, fluency, pace and filler words. The strength is continuity rather than one test morning. It only measures people who use it.
How often should a company reassess employee English levels?
Quarterly is the practical cadence. Movement between CEFR levels takes months, so more frequent reviews mostly capture noise, while annual ones let a stalled programme burn budget unnoticed. With continuous assessment, the quarterly review becomes reading a dashboard rather than organising a testing round.
Which CEFR level do employees need for business English?
It depends on the role. B2 is the common working minimum for jobs with independent external communication, such as sales calls or client meetings. C1 suits roles built on persuasion and negotiation. Internal roles with occasional written English can operate at B1.
Book a Lucida for Business demo
Lucida is an AI speaking coach app for iOS and Android, and Lucida for Business adds team dashboards with a CEFR distribution, per-employee assessments and speaking scores tracked over time. Talk to us and see the dashboard.
