The ROI of Business English: an honest worked example

Calculating the ROI of Business English: why the question is fair

Sooner or later, someone from senior management or the finance team asks: “What do we actually get out of this?” With language training, this question is often seen as awkward, because language seems hard to measure. I think that is a mistake. Language training is an investment, and investments can and should be measured in figures.

At the same time, an ROI calculation is only as good as its assumptions. I have seen too many presentations promoting language training with three-digit returns because the most important costs were simply missing. Figures like that do not survive critical questioning, and in the end they damage the very budget they were meant to protect.

In this article, I therefore present a transparent, deliberately conservative worked example. It is based on the methodology I described in two articles for Pearson, on measuring with the GSE and KPIs and on translating language data into business value. Every calculation starts from a clean baseline, such as the one a language audit provides.

The ROI formula for language training

Following the Phillips methodology, the ROI of language training is calculated as benefits in euros minus costs in euros, divided by the costs, multiplied by 100. The benefit-cost ratio (BCR) also shows whether the investment has paid for itself; a BCR of 1.0 means break-even. The figure is only credible if it is backed by data on reaction, learning, behaviour and results.

For training and development, the ROI methodology developed by Jack Phillips has become established. It adds a fifth level, financial return, to Kirkpatrick’s well-known four-level model. The formula is simple (Training Industry):

ROI (%) = (benefits in euros minus costs in euros) / costs in euros × 100

The benefit-cost ratio (BCR) is often given as well: benefits divided by costs. A BCR of 1.0 means that the investment has exactly paid for itself.

The five levels applied to language training

One rule applies: level 5 is worthless without levels 1 to 4. Without learning and behaviour data, you cannot calculate a credible ROI. How this data is collected is described in the article on the Global Scale of English.

Capturing the full costs of language training

The most common mistake in ROI calculations is an incomplete cost side. In my second Pearson article, I pointed out that just two hours of lessons a week add up to more than 100 hours per person per year, not counting self-study. This time is working time, and it costs money. Course fees are often only a minority of the real costs.

Checklist: these costs belong in the calculation

• Direct costs: course fees, learning platform, materials
• Diagnostics: placement, interim and final tests, language audit
• Learning time in class: lesson hours × number of participants × internal hourly rate
• Self-study time: where it takes place during working hours
• Coordination: time spent by HR, L&D and managers on planning and reporting
• Optional: certificate exams, travel costs for in-person formats (usually zero for online courses)

A good provider should keep the coordination effort as low as possible. This is not a side issue: in larger programmes, internal organisation can noticeably add to the course fees. Read more about the budget side in the article on planning your language training budget.

Measuring the benefits of Business English: hard and soft effects

The benefits of Business English can be measured mainly through hard effects: less language-related rework, lower translation and interpreting costs, faster quotations and relief for colleagues with strong language skills. These can be converted into euros via hours, invoices or conversion rates. Soft effects such as confidence, motivation and retention are reported but usually not monetised.

The soft effects are not an afterthought. In a survey of 200 SME decision-makers published by HR WORKS, 65% saw a lack of English skills as a threat to competitiveness. Effects like these belong in every report, just not in the euro column if they cannot be quantified properly. How to train specific skills such as chairing English-language meetings is shown in the article on leading meetings in English.

Worked ROI example for Business English (illustrative example)

Important note: the following calculation is a constructed, illustrative example. All figures are assumptions, not data from a real customer and not a price quotation. Its purpose is to make the method easy to follow. Your own figures will come from your quotation, your hourly rates and your measurement data.

Starting point

• 12 employees from sales and customer service at an industrial company (example)
• 3 small groups of 4 people each, 2 lessons of 45 minutes per week, 40 weeks
• 80 lessons per person, i.e. 60 clock hours of teaching, plus 20 hours of self-study during working hours
• internal hourly rate (full cost) assumed at €40, €50 for HR coordination

Costs in year 1

Benefits in year 1 (after attribution)

Result

• Year 1: ROI = (€54,877 minus €55,900) / €55,900 × 100 = minus 1.8%, BCR 0.98. The investment almost paid for itself in the first year.
• Year 2: a maintenance format at half the volume, costs €27,950. Benefits conservatively set at 90% of the previous year: €49,389.
• Over two years: benefits €104,266, costs €83,850. ROI = 24.3%, BCR 1.24.

This result may look less spectacular than some advertising promises. That is its strength: it survives the questions of any financial controller.

What tips the ROI of language training: a sensitivity analysis

Two assumptions tip the ROI most. If learning time is left out, it jumps in our example from minus 1.8% to plus 213.6%, a methodologically wrong figure. If the course misses the actual need because there was no needs analysis and the benefits are halved, it falls to minus 50.9%. So the biggest lever is effectiveness, not the course fee.

A single figure is dangerous. It is more meaningful to look at how sensitive the result is to changed assumptions. For our example, year 1 looks like this:

What I conclude from this

First: leaving out learning time produces fantasy figures. The second row is exactly the kind of calculation I see most often in presentations. Second, and more important to me: the biggest lever is not the course fee but effectiveness. In our example, the fees account for only around a quarter of the costs. Training that misses the need above all wastes employees’ working time. A thorough language needs analysis and small groups of four, at most six people are not cost drivers but return drivers.

Isolating effects: how to make your ROI calculation credible

Whether an order was won because of English training or because of a new product can never be said with certainty. The Phillips methodology therefore stresses that the effect of training must be separated from other influences. In practice, three approaches have proved their worth: estimates with a confidence discount, a comparison group that starts later, and a comparison with the long-term trend.

1. Estimate with a confidence discount: managers estimate the share attributable to the training and state how confident they are. The two values are multiplied, as with the orders in our worked example.
2. Comparison group: one team starts a quarter later. The difference in the metrics gives an indication of the training effect.
3. Trend comparison: metrics before and after the training are compared with the long-term trend.

Independent measurement of learning is just as important. If the training provider assesses its own results, the calculation is open to challenge. A standardised placement test at the start and at the end provides the basis. How to present the results afterwards is shown in the article on language training reporting for HR. For a deeper dive in Polish, see the article KPIs for company language courses (in Polish).

Typical objections from finance and good answers

Finance teams almost always raise four objections: the orders would have come anyway, hours saved are not real money, the learning success has not been proven, and AI translation is enough. Good answers are partial attribution with a confidence discount, documented use of the time freed up, independent measurement with a baseline and re-test, and a clear distinction between where AI helps and where it does not.

“The orders would have come without the training anyway.”

Possibly, and that is why we only count them partially and with a confidence discount. In our example, only €18,000 of the €60,000 contribution margin goes into the calculation. If you prefer, you can strike this item entirely and show that over two years the other effects alone cover a large part of the costs.

“The hours saved are not real money.”

A fair objection, as long as the time freed up is not used productively. I recommend clarifying with managers what the time is actually used for and documenting this in the report.

“How do we know people have really improved?”

From independent measurement at the start and at the end, supplemented by observations in the workplace. That is what the baseline and re-test are for. The article on measuring language training success also describes the general approach.

“Why not just use AI translation?”

For routine texts, it is a good complement and should be taken into account in your strategy. In negotiations, customer meetings and spontaneous phone calls, it does not replace people’s language skills. There is more on this in the article on language strategy for internationalisation.

From ROI calculation to budget decision

To turn an ROI calculation into a budget decision, prepare it in advance as a forecast, keep it conservative and look at two years rather than one. Soft effects such as motivation, retention and confidence belong in the report even without a euro value. That way, the calculation becomes an offer to discuss with senior management that stands up to questioning.

For it to work, I recommend four things:

• Calculate before, not just after: a forecast with stated assumptions makes expectations explicit and can be checked later.
• Stay conservative: better a smaller, robust ROI than a big number that falls apart at the first question.
• Think in years: language skills have an effect over years. A two-year period is often more realistic than one.
• Report soft effects: motivation, retention and confidence belong in the report, even if they do not appear in the euro column.

If you want to set up your own programme this way, the choices® methodology team supports you with baseline, KPIs and measurement plan as part of a language needs and level analysis. The audit is a paid service with no obligation to book a course afterwards. Implementation takes place in Business English courses from A1, as one-to-one training or in small groups, embedded in our language courses for companies, from €40 net plus VAT per 45-minute lesson. For a no-obligation conversation, reach us via our contact page.

Sources

1. pearson.com: gse roi business english measurable advantage 11 25
2. pearson.com: gse roi pt 2 language data and business value 12 25
3. trainingindustry.com: phillips roi methodology
4. kirkpatrickpartners.com: the kirkpatrick model
5. hrworks.de: sprachbarriere englisch deutscher mittelstand verliert den anschluss
6. choices.com.pl: jak wyznaczac kpi dla kursow jezykowych w firmie

Frequently asked questions

Łukasz Pakuła

Head of Studies at choices®. Adjunct Assistant Professor at the University of Maryland Global Campus, lecturer at HM Business School Munich and Kempten University of Applied Sciences. Doctoral studies in applied linguistics, publications with Springer and Equinox, among others. More about us