Language audit: 10 mistakes to avoid

What is a language audit, and why does it so often fail?

A language audit systematically records which language skills a company needs, which it already has and how to close the gap. It combines needs analysis, placement testing and goal planning into a reliable basis for decisions on training, recruitment and budget. Our page on the language audit for companies describes the approach in detail.

In theory, it sounds simple. In practice, a surprising amount goes wrong. Not because the people involved are careless, but because an audit has many interfaces: senior management, HR, line managers, the works council (Betriebsrat), employees and an external provider. Every interface is a potential source of error.

The need for this kind of analysis is considerable. In a survey of 200 decision-makers from German SMEs (the Mittelstand) published by HR WORKS (in German), 74% saw insufficient English as an obstacle to collaboration in international teams, and two thirds felt their competitiveness was at risk. All the more frustrating when an audit fails because of avoidable mistakes. We have collected the ten most common, grouped by the phases of planning, measurement, communication and evaluation.

Mistakes when planning a language audit

Three mistakes typically happen at the planning stage: the audit’s goal is unclear, senior management is not involved, and only English is considered. The results: wrong or excess data, no implementation budget and overlooked gaps in German or customer languages. The fix is a clearly stated goal, a sponsor from the leadership team and a complete language needs analysis.

Mistake 1: the goal is unclear

“Let’s see where we stand” is not a goal. An audit for forming training groups needs different data from one preparing a market entry or evaluating a corporate language policy. Without a clear goal, too much data, or the wrong data, gets collected.

Better: state in one or two sentences beforehand which decision the audit is meant to prepare, for example: “We want to know which twelve people in sales need to be able to negotiate confidently in English by the end of the year, and how far they are from that.”

Mistake 2: senior management is not on board

If HR runs the audit alone, the link to business goals is often missing. And once the results are in, there is no budget for implementation.

Better: before you start, bring in someone from senior management or divisional leadership as a sponsor. They should help define the goals and sign off the results.

Mistake 3: only English is considered

In Germany, Austria and Switzerland the biggest bottleneck is often not English at all, but the German of international staff, or a customer language such as French or Polish. An audit that only covers English misses these gaps.

Better: in the language needs analysis, record every language that plays a role in day-to-day work, then prioritise. For new international colleagues it is also worth looking at onboarding international professionals.

Mistakes when measuring language levels

Three mistakes occur during measurement: self-assessment is used instead of a test, speaking is not measured, and the training provider assesses itself. A standardised placement test plus a short 15- to 20-minute spoken interview gives reliable results. Baseline and final measurements should be independent of the provider, for example an external, standardised test.

Mistake 4: self-assessment instead of a test

A questionnaire asking “How good is your English: basic, good, very good, fluent?” is quick to send out. But the results are hardly comparable. Some people underrate themselves out of perfectionism; others overrate themselves because they understand a lot but speak little.

Better: use self-assessment as a supplement, not a substitute. For reliable results you need a standardised placement test.

Mistake 5: speaking is not measured

Many online tests mainly cover reading, listening, grammar and vocabulary. At work, though, speaking is often the biggest weakness. The Germany page of the EF English Proficiency Index shows a clear gap: 623 points for reading and 609 for listening, but only 521 for speaking. If you only test passive skills, you form groups based on a flattering picture.

Better: add a short 15- to 20-minute spoken interview with a qualified teacher to the test.

Mistake 6: the training provider assesses itself

If the same provider places learners, teaches them and finally certifies their progress, there is a conflict of interest. The results may well be accurate, but they are hard to defend to controlling and senior management. In his article for Pearson Languages, Łukasz Pakuła, Head of Studies at choices®, therefore describes independent, GSE-based measurement as an essential component of credible data.

Better: use a standardised, external test for the baseline and final measurement. For English, a test aligned to the Global Scale of English is a good choice.

Mistakes in communication, data protection and works council involvement

Two mistakes put participation and the timeline at risk: employees feel monitored, and the works council and data protection are involved too late. Explain in advance what the audit is for and who sees the results. Present the goal, process, data scope, access rights and deletion periods to the works council before the start, because personnel questionnaires require its consent.

Mistake 7: employees feel monitored

A language test sent out without explanation causes worry: will the result show up in my appraisal? Should I fear for my job? These fears lead to low participation and distorted results.

Better: explain clearly in advance what the audit is for, who sees the results and that it is about support, not assessment. A short message from senior management or divisional leadership often carries more weight here than an email from HR.

Mistake 8: the works council and data protection come in too late

If the works council is only informed once the test links have gone out, trouble is guaranteed. Under Section 94 of the Works Constitution Act (BetrVG) (in German), personnel questionnaires and general assessment principles require the works council’s consent. Whether, and to what extent, a language test or questionnaire falls under this depends on the individual case. Your data protection officer should also be involved early on.

Better: present the goal, process, data scope, access rights and deletion periods before the start. A works council that is involved early often becomes an advocate, especially when it comes to fair learning time for everyone. This information is not a substitute for legal advice.

Mistakes when evaluating language audit results

Two mistakes threaten the evaluation: results without target levels, and a report that disappears into a drawer. A current status such as seven people at B1 only becomes meaningful once a target level has been set for each role. The results only have an impact when they become the baseline for ongoing reporting with fixed interim and final measurements.

Mistake 9: results without target levels

An audit that only shows the current status provides a snapshot with no direction. “Seven people are at B1” only becomes useful information once it is clear whether their roles require B1 or B2.

Better: derive a target level for each role from its communication situations and set current and target levels side by side. Our article on language requirements in job profiles shows how to define requirements properly.

Mistake 10: the report disappears into a drawer

Perhaps the most expensive mistake: the audit is completed, presented and then not used any further. The data goes out of date, and at the next budget meeting everything starts from scratch.

Better: make the audit results the baseline for ongoing reporting in language training. Schedule interim and final measurements at the kick-off and link the results to business goals, if necessary with an ROI calculation.

Overview: language audit mistakes, consequences and solutions

The ten most common language audit mistakes fall into four phases: planning (unclear goal, no sponsor, English only), measurement (self-assessment only, speaking not measured, provider assesses itself), communication (feeling monitored, works council too late) and evaluation (no target levels, report in a drawer). The table shows consequences and solutions and works as a checklist for project planning.

Running a language audit properly: process and timeline

A well-planned language audit for a team of 20 to 50 people takes about four to six weeks. It begins with a kick-off and agreement with the works council and data protection, moves through interviews, questionnaires and placement to the evaluation, and ends with the presentation of results and the transition to training. The process in detail:

1. Kick-off (week 1): clarify the goal, target group, languages, sponsor and responsibilities.
2. Agreement (weeks 1 to 2): inform the works council and data protection, prepare communication to employees.
3. Interviews (week 2): talks with managers about critical communication situations.
4. Questionnaire (weeks 2 to 3): language use, self-assessment, learning experience and available time.
5. Placement (weeks 3 to 4): standardised test plus a short spoken interview.
6. Evaluation (weeks 4 to 5): current and target levels per role, priorities, proposed groups.
7. Presentation of results (weeks 5 to 6): report to the sponsor and HR, decision on measures.
8. Transition to training: courses start, KPIs and measurement dates are set.

This structure is a guideline. Larger organisations with several sites in Germany, Austria and Switzerland are well advised to work in waves, starting with a pilot area.

What a good audit report contains

• a one-page summary for senior management
• the most important communication situations per department, prioritised by criticality
• current and target levels per role, broken down by skill
• concrete recommendations on formats: small groups of four to six people, one-to-one training for key roles
• a measurement plan with dates for interim and final tests and a few clearly defined KPIs
• open issues and risks, such as a lack of learning time or planned restructuring

A report with these elements is not a box-ticking exercise but a working document. It should come back to the table with every later report.

Language audit checklist: tick these off before you start

Before a language audit starts, ten points should be settled: a written goal, a sponsor from the leadership, all relevant languages, a test with a speaking component, an independent baseline, transparent communication, an informed works council and data protection officer, target levels per role, fixed measurement dates and clear responsibility for reporting.

Before you launch the audit, you should be able to answer “yes” to each of the following points.

• The goal of the audit is stated in one or two sentences.
• Someone from senior management or divisional leadership is sponsoring the project.
• All relevant languages are covered, not just English.
• There is a standardised test that also covers speaking.
• Baseline and final measurements are independent of the training provider.
• Employees know what the audit is for and who will see the results.
• The works council and data protection officer have been informed.
• Target levels per role are derived from communication situations.
• Dates for interim and final measurements are fixed.
• It is clear who will feed the results into ongoing reporting.

An older but broad international finding also shows why systematic measurement matters: according to Personalwirtschaft (in German), a study by Cambridge English and the QS Institute covering 5,373 companies in 38 countries found that around 40% of employees had gaps in their English. Such gaps are nothing to be ashamed of, but you can only close them if you know about them.

choices® carries out language audits online for companies of every size, with qualified teachers from 10 countries. The audit is a paid service with no obligation to book a course afterwards, and we sign an NDA on request. Read more about our approach on our page on the language assessment for companies. For implementation, there are Business English courses, online German courses and many other languages, with all options listed under language courses for companies. If you would like to talk to us directly, please use our contact page.

Sources

1. hrworks.de: sprachbarriere englisch deutscher mittelstand verliert den anschluss
2. ef.de: germany
3. pearson.com: gse roi business english measurable advantage 11 25
4. gesetze-im-internet.de: 94
5. personalwirtschaft.de: im job fehlt es oft an sprach know how

Frequently asked questions