How to Assess English for Real Workplace Communication
Author: Ksenia Izotova
Last updated: September 2026
Last updated: September 2026
How can HR assess English for real work?
Use a general language assessment to establish a common reference, then add short tasks that reflect the employee's role. Ask a developer to explain a technical choice, a QA engineer to write a bug report or a manager to handle a feedback conversation. Score each task against clear criteria such as task completion, structure, interaction, audience fit and language control. Use more than one sample for decisions that affect placement or development. The result should tell HR what the employee can do, where communication breaks down and which practice would help. A single CEFR label does not provide that detail.
Many companies discover the gap after a surprising meeting. An employee has a solid test result and understands the discussion, yet struggles to enter the conversation, answer a challenge or explain a risk without a long detour.
The test may be accurate. It answered a broader question.
Many companies discover the gap after a surprising meeting. An employee has a solid test result and understands the discussion, yet struggles to enter the conversation, answer a challenge or explain a risk without a long detour.
The test may be accurate. It answered a broader question.
Assessment design should match its purpose. Use qualified assessors, inform employees how results will be used and follow the organisation’s privacy and employment policies.
Define the decision before the assessment
Assessment design changes with its purpose. HR may need to:
- place employees in learning groups;
- identify communication priorities for a team;
- plan support for a new client-facing role;
- measure progress after a programme;
- give an employee a development plan.
These decisions do not require the same amount or type of evidence. A short placement check can support group formation. A high-stakes employment decision needs a more formal, validated process and appropriate professional and legal oversight.
Write down what the result will and will not be used for. Tell participants before they begin. This reduces anxiety and prevents a development assessment from quietly becoming an undisclosed performance screen.
Write down what the result will and will not be used for. Tell participants before they begin. This reduces anxiety and prevents a development assessment from quietly becoming an undisclosed performance screen.
Use CEFR as a reference rather than a job description
The Common European Framework of Reference for Languages gives organisations a shared language for describing proficiency. The CEFR Companion Volume covers reception, production, interaction and mediation, with illustrative "can do" descriptors across levels.
This is broader than a grammar test and useful for programme planning. It still cannot know that one B2 employee leads retrospectives, while another writes technical documentation and rarely joins external calls.
A workplace profile therefore has two parts:
This is broader than a grammar test and useful for programme planning. It still cannot know that one B2 employee leads retrospectives, while another writes technical documentation and rarely joins external calls.
A workplace profile therefore has two parts:
- a general description of language ability;
- evidence from the communication tasks required by the role.
Avoid treating the second part as a replacement CEFR level. “B2 for stand-ups” sounds precise and can confuse two different things. Report the CEFR reference and the task result separately.
Map the communication before building the task
Ask four practical questions:
- Who is the employee communicating with?
- What should the other person understand, decide or do?
- What information or uncertainty must the employee manage?
- What usually goes wrong?
Suppose a DevOps engineer joins incident calls. The work does not require a speech about cloud infrastructure. It requires short updates under pressure, careful use of hypotheses, clear handovers and responses to questions from people with different levels of technical knowledge.
That map can produce a task such as:
“You have confirmed elevated error rates after a deployment. The cause is still unknown. Give a 90-second update to Engineering and Support, then answer two questions.”
The task creates a reason to communicate. It also allows the assessor to see whether the participant separates facts from possibilities.
That map can produce a task such as:
“You have confirmed elevated error rates after a deployment. The cause is still unknown. Give a 90-second update to Engineering and Support, then answer two questions.”
The task creates a reason to communicate. It also allows the assessor to see whether the participant separates facts from possibilities.
Choose evidence that resembles the work
Assessment materials should use fictional or sanitised company contexts. Employees should not have to reveal client data or sensitive incidents to demonstrate their English.
Include the difficult second turn
A prepared monologue often overstates workplace ability. Meetings change when someone interrupts or asks, “How sure are you?”
After the first response, add one or two realistic prompts:
After the first response, add one or two realistic prompts:
- “What evidence supports that?”
- “Can we still release on Friday?”
- “Why was this not raised earlier?”
- “Could you explain that without the technical detail?”
The assessor can then observe clarification, repair, turn-taking and control under modest pressure. This is often where a participant’s real development need becomes visible.
Keep the interaction consistent enough for fair comparison. Different follow-up questions are fine when they test the same demand and stay within the task brief.
Keep the interaction consistent enough for fair comparison. Different follow-up questions are fine when they test the same demand and stay within the task brief.
Use a rubric people can understand
An assessor’s overall impression is difficult to explain or repeat. A short analytic rubric makes the judgement more useful.
Define what each performance band means and train assessors with sample responses. If two assessors use a criterion differently, the report may create false confidence.
Cambridge English sets out a systematic approach to assessment development in its Principles of Good Practice. A corporate workplace task will usually be smaller and more local than a formal examination. The same discipline still matters: the task, scoring and intended interpretation should fit the decision.
Cambridge English sets out a systematic approach to assessment development in its Principles of Good Practice. A corporate workplace task will usually be smaller and more local than a formal examination. The same discipline still matters: the task, scoring and intended interpretation should fit the decision.
Do not score accent as a proxy for competence
Pronunciation matters when it affects intelligibility. Accent is not the same thing. An employee can speak with a noticeable accent and communicate precisely; another may sound fluent and leave the decision unclear.
Assessors need a defined question: can the intended listener understand the message without unreasonable effort? Note the specific feature when it interferes, such as stress on a critical number or a sound contrast that changes a technical term. Avoid broad comments such as “needs a more native accent.” The CEFR itself does not define C2 as native-speaker competence.
Assessors need a defined question: can the intended listener understand the message without unreasonable effort? Note the specific feature when it interferes, such as stress on a critical number or a sound contrast that changes a technical term. Avoid broad comments such as “needs a more native accent.” The CEFR itself does not define C2 as native-speaker competence.
Use more than one sample for important decisions
Performance changes with topic familiarity, fatigue, preparation and the other speaker. One five-minute task can be enough for an initial programme conversation. It is weak evidence for a consequential decision about an employee.
Where the stakes or investment are higher, combine:
Where the stakes or investment are higher, combine:
- a general assessment;
- two or more role-relevant tasks;
- speaking and writing where both matter;
- structured self-assessment;
- manager input limited to observed situations.
The sources can disagree. That disagreement is informative. A participant may perform well in a simulation and avoid speaking in real meetings because the team climate feels unsafe. Another may communicate effectively in a familiar role despite uneven grammar.
Report a profile HR can use
“Intermediate” gives the programme designer little direction. A useful report might say:
The participant communicates routine technical information clearly with engineers. In a client explanation, the main issue appears after several details and the requested decision remains implicit. Recommended practice: lead with impact, separate confirmed facts from assumptions, and close with a specific question or next step.
For group planning, aggregate the findings:
The participant communicates routine technical information clearly with engineers. In a client explanation, the main issue appears after several details and the requested decision remains implicit. Recommended practice: lead with impact, separate confirmed facts from assumptions, and close with a specific question or next step.
For group planning, aggregate the findings:
Do not give managers detailed individual recordings or comments unless the purpose, consent and policy support that access. HR usually needs enough data to design and evaluate the programme, not a collection of personal mistakes.
From the needs questionnaire to a placement recommendation
The needs questionnaire and the language assessment should remain separate. One describes the work. The other describes current language ability.
A short questionnaire can follow this route:
A short questionnaire can follow this route:
- Ask about the person’s role and where English appears in their work.
- Ask whether the current need is specific or mainly general development.
- Let the person select up to three relevant situations, such as meetings, writing, presentations or client communication.
- If the need is specific, ask for the exact tasks and their frequency.
- Ask what currently gets in the way.
The answer “meetings” is useful context, but it does not yet define a specialised course. “Run two stand-ups each week and respond when colleagues challenge the estimate” gives the programme designer a task, a frequency and an interaction to assess.
The placement process then adds language evidence. This may include a written or online test and, where speaking matters to the decision, a short oral interview. The assessor records the profile, recommends a level and identifies suitable group options. Employees should know how the result will be used and who can see it.
The placement process then adds language evidence. This may include a written or online test and, where speaking matters to the decision, a short oral interview. The assessor records the profile, recommends a level and identifies suitable group options. Employees should know how the result will be used and who can see it.
Warning signs in a workplace assessment
- The test has no stated purpose beyond “check everyone's English.”
- Every role completes the same generic speaking prompt.
- The final result is a precise number with no task description.
- Fluency is judged mainly by speed or accent.
- Employees do not know who will see their results.
- The provider cannot explain the scoring criteria.
- A single short task drives high-stakes decisions.
- The report gives labels without development recommendations.
Any one issue deserves a question. Several together suggest that the assessment may be easier to administer than to use responsibly.
Frequently asked questions
Is an online level test enough for corporate group placement?
It can provide a useful first reference. Add speaking or writing evidence where those skills matter, and check the roles and communication tasks before finalising groups.
Should technical knowledge be scored?
Only to the extent needed to complete the communication task. Give participants the necessary facts so the assessment does not become an engineering exam. Score how they communicate the information, unless technical judgement is an explicitly separate assessment.
Can managers assess their own employees' English?
Managers can describe observed workplace behaviour. A trained language assessor should interpret language performance when a formal judgement is needed. Combining the perspectives is often more useful than asking either person to answer every question.
How long should a workplace English assessment take?
There is no universal duration. The task should be long enough to produce the evidence needed for the decision. Use shorter screening for low-stakes placement and broader sampling for high-stakes conclusions.
Should employees receive their results?
They should receive a clear explanation of their own development findings and recommendations, subject to the agreed assessment and privacy policy. Decide this before testing and communicate it openly.
Assess what people are expected to do
UnifyHub's team English assessment combines language reference points with roles and workplace communication needs. The findings can shape corporate English training for IT teams and role-specific programmes for software engineers, QA engineers and engineering managers.
Choose one recurring task before designing the assessment. If the task cannot explain what the employee is expected to accomplish with another person, it is probably still too broad.
Choose one recurring task before designing the assessment. If the task cannot explain what the employee is expected to accomplish with another person, it is probably still too broad.
Author: Ksenia Izotova
Last updated: September 2026
Last updated: September 2026