Build a Bilingual Customer Segmentation Model
Segment your market in Arabic and English without losing nuance in either language
Customer data arrives in mixed languages and your segmentation breaks when cultural categories do not translate directly. Running analysis only in English misses how Arabic speakers actually describe their needs and preferences.
That is the gap this drop closes. It is a data & analysis skill built for Kimi, and it takes about ten minutes to set up the first time. After that it runs in under a minute.
Who should use it
Business analysts and marketers working with Arabic and English customer data
How it works
The skill file does five things, in this order.
1. Map category labels in both languages first. List your segment names in English, then write the Arabic equivalent that your customers would actually use
2. Test each segment definition for translation drift. Give Kimi a segment definition in English and ask it to identify what gets lost or added when expressed in Arabic
3. Generate bilingual survey questions. Have Kimi write questions that probe the same dimension in both languages while respecting different conversational norms
4. Run parallel analysis on mixed data. Process Arabic responses in Arabic and English responses in English, then ask Kimi to reconcile findings into unified segments
5. Document terms that do not translate. Keep a glossary of segment descriptors where one language captures a distinction the other misses
What comes back
A segmentation framework that works in both languages with clear notes on where each language adds unique insight
The mistake to avoid
Income brackets translate numerically but spending attitudes do not. What counts as premium or value differs culturally, so test your segment boundaries with native speakers before finalizing.
Running it
Paste the prompt block into a new Kimi chat and attach your files in the same message. Kimi's long context means you should upload everything at once rather than in batches, splitting the input is what causes it to lose cross-references between documents.
Where this fits
On its own, one skill saves an hour a week. The compounding happens when three or four of them run in sequence on the same input, the same transcript that produces a scope of work also produces the follow-up email and the project brief. That is the point at which it stops being a prompt and starts being an internal tool. If you want that wired into the systems your team already uses, that is the work 67 Digital does.
In the file
I am segmenting customers in the {industry} market across the UAE.
I have these initial segment definitions in English:
{segment_definitions}
For each segment:
1. Provide the Arabic term or phrase customers would naturally use to describe themselves in this category
2. Identify any cultural or behavioral nuances that the English definition misses when applied to Arabic-speaking customers
3. Identify any aspects that the Arabic framing reveals that the English version overlooks
4. Flag where income, family structure, or social expectations create different segment boundaries between the two language groups
Then suggest 3 survey questions for each segment that work in both languages. Write each question in English and Arabic, ensuring both versions probe the same underlying dimension while respecting how each language group discusses {industry} preferences.
Finally, create a reconciliation table showing:
- Segment name in English
- Segment name in Arabic
- Core defining criteria that apply to both
- Criteria unique to English-speaking customers
- Criteria unique to Arabic-speaking customers
Format the output so I can use it directly in my segmentation model documentation.