> ## Documentation Index
> Fetch the complete documentation index at: https://docs.kehillahq.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Importing members and households

> Bring your existing directory across from a spreadsheet.

If your congregation already keeps its directory in a spreadsheet, and most do, you can bring it across in one pass rather than typing it in.

Go to **Members** and choose **Import**.

## Before you start

You do not need a specially formatted file. The import reads your spreadsheet as it is and works out what the columns mean. But a few minutes of tidying makes the result better:

* **One person per row** works best. Rows holding a couple, such as "David and Miriam Rosen", are handled, but check the result.
* **Make sure households are identifiable.** A household name column is ideal. Without one, people are grouped by shared surname and address.
* **Clean up obvious duplicates** in the spreadsheet first. Easier there than afterwards.

CSV and Excel files both work.

## How the import works

<Steps>
  <Step title="Upload your file">
    The file is read and analysed. Nothing is saved yet.
  </Step>

  <Step title="Check the column mapping">
    You are shown what each of your columns has been matched to, for example "Email Address" matched to Email. Correct anything wrong, and mark columns you want ignored.

    Take your time here. This step is where import quality is decided.
  </Step>

  <Step title="Answer any questions">
    Where a column is genuinely ambiguous, you are asked directly rather than guessed at.
  </Step>

  <Step title="Review households and duplicates">
    You see how rows have been grouped into households, and which records look like people already in your system.

    For each possible duplicate, choose what to do: **skip** it, **update** the existing record with the new information, or **create** a separate record because they really are two different people.
  </Step>

  <Step title="Commit">
    Records are created. You get a per-row report of what actually happened.
  </Step>
</Steps>

<Note>
  The report tells you the truth, including failures. If eight rows did not import, it says eight rows did not import and why. A count that looks lower than you expected is information, not a display bug.
</Note>

## What happens to households

Households are found or created by name. If a household in your file already exists, people are added to the existing one rather than a duplicate being made.

Where a household has no primary contact, one is inferred from the people in it. Check these afterwards and adjust where the guess is not right.

## Joined dates

If your spreadsheet has a date the family joined, map it. It decides which dues years each household appears on, and it is much easier to bring across now than to reconstruct later. See [Households that join mid-year](/finances/mid-year-joins).

## After the import

Give yourself half an hour to check the result:

* Spot-check a few households, especially any that were couples on one row.
* Look at the individuals view filtered to records with no email address. Those people will not receive yahrzeit reminders.
* Check the family roles, Adult and Child, if you use the Fair Share dues model. Fair Share brackets count active adults.
* Confirm the primary contacts.

## Importing yahrzeits

Yahrzeit records import separately, because they involve Hebrew dates. Do your members first, so yahrzeits can be linked to the right people. See [Importing yahrzeits](/yahrzeits/import).

<Tip>
  Import into a quiet moment, not ten minutes before a board meeting. It is not slow, but you will want unhurried time to review what came in.
</Tip>