CardKeeper: Business Cards · How it compares
Built to remember people, not to collect them.
Most tools that touch business cards are built around a platform, a team or a sales pipeline. CardKeeper is built for one person who meets people and wants to remember them, privately.
The usual choices
| Alternative | What it does well | Where CardKeeper is different |
|---|---|---|
| The iPhone camera's built-in text recognition | Free and built in; turns a single number or email into a contact. | A one-off copy of text. CardKeeper keeps the card image, where you met, duplicate checks, a review step and your Mac. |
| Dedicated card-scanner apps | Mature scanning; some export to sales software. | Many ask for an account and keep your contacts on their servers. CardKeeper has no account and keeps contacts in your own iCloud. |
| Digital business-card platforms | Sharing your own card; capturing leads for teams. | Built around their card and their subscription. CardKeeper is for the cards you receive and the people behind them. |
| A sales CRM | Pipelines, teams and reporting. | Heavy for remembering people. CardKeeper needs no setup and has no pipeline. |
| Contacts and a notes app | Free and familiar. | Loses the card and the context. CardKeeper keeps both with every contact. |
Real-card reading test
On October 6, 2026 we scanned 11 real business cards once each in CardKeeper (beta build 279, with AI card reading allowed) and in a competing card-scanner app (free tier). We saved each first result without corrections and checked every printed detail against the card itself. The test exposed phone-number formatting problems in CardKeeper, so we fixed them and rescanned the same 11 cards the same day on build 289. The competing app was scanned once.
The cards came from Japan, the UK, Italy, Costa Rica, the Turks and Caicos Islands and the US. They included all-caps print, vertical layouts, logo badges, extensions, toll-free numbers, two hotels on one card and a QR code.
| 11 real cards | CardKeeper, first test | CardKeeper, after fixes | Competing app |
|---|---|---|---|
| Cards needing no correction | 2 of 11 | 9 of 11 | 0 of 11 |
| All printed details | 60 of 70 (86%) | 66 of 70 (94%) | 31 of 70 (44%) |
| Names | 10 of 10 | 10 of 10 | 7 of 10 |
| Companies | 8 of 9 | 8 of 9 | 1 of 9 |
| Job titles | 9 of 9 | 9 of 9 | 6 of 9 |
| Email addresses | 11 of 11 | 11 of 11 | 4 of 11 |
| Websites | 6 of 7 | 6 of 7 | 0 of 7 |
| Addresses | 4 of 6 | 4 of 6 | 0 of 6 |
| Phone numbers | 12 of 18 | 18 of 18 | 13 of 18 |
| Details added that are not on the card | 3 | 0 | 5 |
What we saw
- Email is where the difference matters most. The competing app produced plausible but wrong addresses, a "9" for a "g", or "travet" for "travel", that would bounce without warning. CardKeeper read all 11 correctly.
- Labels and badges are not people. The competing app named one contact "Toll-Free Us" and turned a hotel's membership badges into a person called "Fine Hotel". CardKeeper left the name empty on a card that printed no person.
- Phone numbers were CardKeeper's weak spot, now fixed. The AI read every number correctly, but CardKeeper's own formatting mishandled the UK "(0)", a "(+)1" number and US toll-free numbers printed on foreign cards. After the fix, all 18 numbers came out right.
- Still not perfect. After the fixes, 9 of the 11 cards needed no correction, against none for the competing app. The other two were a hotel card that prints two properties, which CardKeeper merged into one entry, and an address that left out the printed region name. That is why CardKeeper highlights uncertain fields and never saves without your review.