A regional hub in Dublin or London often has one playbook, written in English, and calls in six or seven languages. The playbook says how to open, qualify, handle a price objection and close. Then each rep applies it in German, Polish, Spanish or French, and the only person who can judge how it sounds is another speaker of that language.
Coaching in English is easy to scale. Coaching in every language is not. This page describes a routine that gives every rep the same practice scenario in their own market language, a weekly review that takes fifteen minutes, and a way for the team lead to see patterns without needing to speak every language.
The problem in practice
- Training lands in English only. Workshops and roleplays run in the common language. Reps then work in another one.
- Quality varies by language and nobody can see it. A manager who does not speak Polish cannot hear whether the Polish reps hold the call structure.
- Calls are expensive to review. Listening to recordings takes time, and in some contexts you may not want to store recordings at all.
- New hires wait. A new German-speaking rep may wait weeks for a colleague to have time for a roleplay.
Step 1: build one scenario, not seven
Pick a situation that occurs in every market. For example:
- A customer says the price is too high.
- A customer wants to cancel.
- A prospect says "send me an email".
Write the scenario once, in plain terms: who the customer is, what they want, what they object to and what a good outcome is. Keep the structure the same across languages, and change only the language and local details such as currency and names. Invented names and details only; no real customer data.
For ideas on scenarios, see sales roleplay exercises for teams and customer service roleplay scenarios.
Step 2: run it in each market language
Each rep runs the scenario by voice, in the language they sell in. With DIRI AI the practice customer speaks that language, and the report comes back after the call. Eleven languages are available: Czech, English, Slovak, Polish, German, Spanish, Portuguese, Italian, French, Romanian and Hungarian.
Fifteen minutes is enough: one call, a few minutes with the report, and one change to try. Longer sessions tend to get skipped.
Step 3: a weekly 15-minute review per rep
Weekly practice only works if there is a short, predictable review. A format that fits:
- The rep picks one moment from the report. The one they want to fix.
- They read the stronger wording the report suggests, in their own language. Then say it aloud twice.
- They run the call once more and compare.
- They write one line in a shared note: the moment, the fix, the date.
The team lead reads the shared notes, not every call. See run weekly call practice sessions for a longer version of the rhythm, and coach sales reps from call reports for what a lead can do with the output.
Step 3b: handle the language gap honestly
A lead who does not speak a language cannot judge its quality. Be explicit about what each person can and cannot assess.
- Structure can be judged in any language: did the rep open clearly, ask a question, handle the objection and agree a next step? The report shows this.
- Naturalness and register need a native speaker. Ask one native colleague per language to review the shared wording once a quarter, and keep a short list of approved phrases per market.
- Local rules and norms come from people who work in that market, and from your compliance contact where regulation applies.
This keeps coaching useful without pretending that one manager can cover every language.
What a manager sees on company plans
DIRI AI company plans include a team results and progress dashboard, shared scenarios for the whole team and a prioritised manager coaching queue. Practically, that means:
- One scenario can be shared with the whole team.
- A lead can see who has practised and read each rep's reports in their company.
- Reports can feed one-to-one coaching.
Reports are coaching aids. Scores must not be the sole basis for decisions about hiring, discipline, pay or promotion; the Trust Center states this and describes who can see what. Treat practice as practice.
Starting small
Team plans start with a pilot request. Individuals can start alone on their own plan, which is a low-risk way to test one scenario in one language before involving the team.
- Week 1. One scenario, one language, two reps.
- Week 2. Add a second language. Compare how reps in each language rate the scenario.
- Week 3. Add the weekly review.
- Week 4. Decide whether to request a team pilot.
You can send a team pilot request. Individual reps can start free. See pricing for what each plan includes. If you are still choosing a tool, the guide to AI training for call centre agents is a good start, and how to choose AI sales roleplay software covers the criteria.
Mistakes to avoid
- Only practising in English. The team works in many languages, so practise in them.
- A different scenario in every language. You then cannot compare reports. Keep it the same.
- Judging language quality from structure alone. Native review is separate.
- Making it a test. Practice should be safe to do badly. Keep reports away from performance ratings.
FAQ
Can the same scenario really work in several languages?
Yes, if it is built around a situation, not a phrase. The goal and objections stay the same and the wording is local.
Do I need to speak every language on my team?
No. You can read structure from the reports and rely on native speakers for naturalness. Be clear about which is which.
How do we start without a procurement process?
Start with a few reps on individual plans, then send a team pilot request if it helps.
Are the reports a performance rating?
No. They are coaching aids for practice.