You come back from a conference with forty business cards and zero usable contacts. They'll sit in a drawer until they're worthless, because nobody types forty cards into a CRM by hand. OCR fixes that: photograph each card, extract the name, title, company, phone, and email, and build a clean contact list you can actually import and use.
Why business cards are a great OCR target
A business card is small, printed, and high-contrast, which is exactly the kind of image OCR handles best. Our engine is Tesseract-based, strong on clean printed text, so a sharp photo of a standard card usually reads accurately. The fields you care about are predictable too: name, job title, company, phone, email, and website. That predictability makes cards far easier to process than messy documents.
The payoff is speed and reliability. A card extracted the day you receive it becomes a contact you'll actually follow up with, rather than a name you can't quite place three weeks later.
What can trip OCR up on a card
Designers love cards that fight legibility. Watch out for:
- Decorative or script fonts that prioritize style over readability.
- Low-contrast text, like grey ink on a pale background or white on metallic.
- Logos and icons mixed into the text, which can produce stray characters.
- Vertical or rotated text that the recognizer reads out of order.
None of these are fatal, but they mean you'll proofread a card that's all flourish more carefully than a plain one.
Step by step: card to contact list
- Photograph the card flat and lit evenly. Lay it on a contrasting surface and shoot straight down to avoid glare on glossy stock.
- Crop to the card edges. Remove the background so the recognizer only sees the card.
- Extract the text. Run it through image to text to get the fields as editable text.
- Split the fields. Separate name, title, company, phone, and email into the columns your contact tool expects.
- Save to a structured list. Drop everything into image to excel so each card becomes a clean row ready to import into a CRM or address book.
Building a clean, importable list
A pile of extracted text is only half the job; contacts need structure. Pulling card data into image to excel lets you assign a column to each field, which is exactly the layout a CRM import expects. One row per person, consistent columns, and your list is ready to upload.
If you collected a whole stack at an event, don't process them one at a time. Photograph every card, then run the set through batch image to text and paste the results into your sheet in a single pass. The capture habits are the same ones used for receipts and expenses, so once you've got the photo technique down it transfers straight across.
Cleaning up the fields
Extracted card text often needs light tidying: a phone number split across lines, a job title that wrapped, an email with a stray character from a nearby logo. Fix these in the spreadsheet before you import, since a malformed email is worse than no email. For longer write-ups about a contact, you can also send notes into a Word document and keep them alongside the structured list.
Accuracy and a privacy note
Always verify emails and phone numbers against the card, character by character, because a single wrong digit means a message that never arrives. Treat the data as personal information too: business cards contain real people's contact details, so handle the list responsibly and don't share it beyond its purpose. Files you upload are processed for the conversion and not retained, but it's still wise to delete working copies once your contacts are safely imported. If you also keep handwritten notes about who you met, our guide on how to digitize handwritten notes covers turning those into searchable text.
Frequently asked questions
How accurate is OCR on business cards?
On standard cards with clear, printed text, it's very accurate, because cards are small, high-contrast, and printed. Decorative fonts, low-contrast color schemes, and logos mixed into text are the cases that need careful proofreading. Always verify emails and phone numbers before importing.
What's the best way to turn cards into a CRM-ready list?
Extract the text, then put it into image to excel with one column per field, name, title, company, phone, email, so each card becomes a row. Most CRMs import that column layout directly.
Can I process a whole stack of cards from an event at once?
Yes. Photograph every card first, then upload the set to batch image to text. You convert them all in one pass and paste the results into your spreadsheet, rather than handling each card separately.
Why are some fields coming out jumbled?
Usually it's design, not the engine. Script fonts, low-contrast ink, rotated text, and logos sitting in the text all confuse OCR. Re-shoot with better lighting and contrast, and expect to tidy the trickier cards by hand in your spreadsheet.
Clear that drawer of cards before the contacts go cold. Start with image to text, then build your list in image to excel.