Likely duplicates when importing
Before anyone is added, Yahiv shows people in your file who look like someone already here — or like another row in the file — and lets you decide: same person or different people.
- 1
Start the import
After checking your files and columns, click “Import everything”. If Yahiv finds likely duplicates, it stops and shows “Check likely duplicates before importing”. Nobody has been added yet.
Start from Settings → Import. - 2
Read each pair
Each pair is shown side by side — “In the file · row …” and “Already in Yahiv …” (or two rows of the file) — with how sure Yahiv is (for example “92% sure”) and why (same phone, same email, same name and city…).
- 3
Decide
For someone already here: “Same person — merge into existing” or “Different people”. For two rows of the file: “Same person — combine” or “Different people”.
- 4
Accept the very likely ones at once
Click “Accept all ≥ 90% as the same person” to answer every very likely pair in one go. Tick “Only unanswered” to see what's left.
- 5
Continue
Click “Continue the import”. Pairs you didn't answer are imported the usual way (matched by old ID, email, phone or exact name).
Tips
- “Merge into existing” only fills in details the person is missing — nothing they have is changed — and their gifts in the file go to that person.
- Two rows with different IDs in the old system (for example Donary accounts 1001 and DNR1002, both “Moshe Katz”) start out as “Different people” — the pair says “Kept apart: they had different IDs in the old system”. “Continue” never merges them; choose “Same person” (or “Accept all ≥ 90%”, which says how many of these it includes) only if they really are one.
- Big files show the 500 most likely pairs.
- “Back to the files” returns to your files without importing.
- The import's review page shows how many were merged, combined and kept apart.
If something isn't working
Two different people were merged.
If they had different IDs in the old system, import the same files again — they are split back out, with their pledges and gifts. Otherwise undo the import from its review page and import again, choosing “Different people” for that pair.
Duplicates got in anyway.
Merge them afterwards with People → Find duplicates (see “Merge two duplicate records”).