A card scanner that works perfectly in London can fall apart in Dubai, Shanghai or Delhi. The cards themselves are the reason: two languages on one piece of card, a reversed side, a right-to-left layout, or a name that exists in two scripts and two spellings. This guide covers what actually goes wrong on multilingual cards and how to capture them without losing half the contact.
Why bilingual cards break ordinary scanners
A scanner does two jobs: recognise the characters, then decide which field each string belongs to. On a monolingual card the second job is mostly layout logic — the biggest text near the top is the name, the string with an @ is the email.
Put two languages on the same card and that logic gets confused. The engine sees two names, two companies and two job titles, and has to work out that they are translations of each other rather than four separate people. Weak scanners either pick one at random or merge them into a single unusable string.
Arabic and the right-to-left problem
Gulf business cards typically carry English on one side and Arabic on the other, or both on one face with the Arabic set right-aligned. Right-to-left text throws off engines that assume reading order runs the other way, which produces reversed strings or numbers detached from their labels.
It also affects phone numbers, which stay left-to-right even inside right-to-left text. The practical advice is to scan both faces as one contact rather than two, and to check the number fields specifically after the first few scans until you trust the output.
Chinese cards, vertical text and two names
Chinese and Japanese cards may set text vertically, and a senior contact will often have a Chinese name and an adopted English one that do not correspond phonetically. Both matter: the English name is what you will type into an email, the Chinese name is what appears on a contract.
Keep both. Store the English name in the primary name field so search works the way your habits expect, and keep the local-script name in the contact rather than discarding it. A scanner that reads 30+ languages, as BizConnect does, will capture the characters rather than dropping the line as noise.
Devanagari, Hindi and the Indian card
Indian business cards mix English with Hindi and regional scripts, often within the same block of text, and frequently include multiple phone numbers with no labels distinguishing office from mobile.
The multi-number problem is the bigger practical issue here. Scan, then take three seconds to confirm which number went into which field, because a mobile filed as a landline breaks every automated follow-up downstream. We looked at this in more depth in the visiting card scanner guide for Indian small businesses.
How to scan a two-language card properly
Treat the card as one contact with two faces. Scan the primary side first, then add the second side to the same record rather than creating a duplicate — otherwise your CRM ends up with two entries for the same person in two scripts, which is worse than having one.
Light matters more on these cards than on plain English ones, because scripts with fine strokes lose detail quickly in shadow. Flat surface, even light, no glare, and fill the frame. If a scan comes back wrong, our notes on fixing blurry scans and bad reads apply just as much to Arabic and Devanagari as to English.
Getting multilingual contacts into a CRM without mangling them
Exports are where non-Latin characters go to die. A CSV saved without UTF-8 encoding turns Arabic and Chinese into question marks, and once the mangled file is imported, the original is gone.
Export to a destination that handles Unicode natively — Google Contacts, Outlook, Salesforce, HubSpot and Zoho all do — rather than round-tripping through an ad-hoc spreadsheet. If a spreadsheet step is unavoidable, our guide to exporting scans to Excel covers keeping the encoding intact.
The bottom line for multilingual networking
If you work across regions, language coverage is not a nice-to-have feature buried in a comparison table — it is the thing that determines whether half your contacts are usable. Check that your scanner handles the scripts you actually meet, scan both faces into one record, verify the number fields, and export somewhere that respects Unicode. BizConnect reads cards in more than 30 languages and is available on Google Play, with a free plan of 50 cards a day to test on your own stack of bilingual cards.
Frequently Asked Questions
Can a card scanner read Arabic business cards?
Yes. BizConnect reads cards in more than 30 languages including Arabic. Scan both faces into a single contact and check the phone number fields, since numbers stay left-to-right inside right-to-left text.
Does a business card scanner work with Chinese characters?
Yes. Keep both the Chinese and the English name on the contact — the English name for search and email, the local-script name for contracts and formal correspondence.
Can it read Hindi and other Indian scripts?
Yes. Devanagari and other Indian scripts are supported. The bigger thing to verify on Indian cards is which of the several printed numbers landed in the mobile field.
Should I scan both sides of a bilingual card?
Yes, into the same contact record. Scanning each side separately creates two entries for one person, which is harder to fix later than it is to avoid now.
Why do non-English characters break after export?
Almost always an encoding problem. Export to a Unicode-safe destination like Google Contacts, Outlook or your CRM directly, or make sure any CSV step is saved as UTF-8.