Prepare your source data

Data source replication

Integration Imports replicate your source data 1-1. A customer record in your source will have an identical row in your CSV file. Similarly, a purchase record in your source will have an identical row in your CSV file.

File format

Each data type is imported from its own CSV file. The column headers for each file are listed in the Schema Reference, along with a downloadable header file per data type. Dates and timestamps should be formatted as ISO-8601, either in UTC (Z) or with an explicit offset, and currency codes as ISO-4217.

This matters most for a purchase's processed_at, which determines the day a purchase is attributed to in reporting. Include the local time offset of the purchase so the timestamp resolves to the correct instant — either convert the value to UTC (Z) or append the explicit offset. A purchase made at 2026-05-11 08:02 ET should be sent as 2026-05-11T12:02:00Z or 2026-05-11T08:02:00-04:00. Do not send the local wall-clock time as if it were UTC (for example, 2026-05-11T08:02:00Z), as this shifts the purchase to the wrong time.

For null datapoints, simply leave the cell blank.

Unique identifiers

Every file is keyed by an id column. We recommend always using a unique, opaque primary key from your source system for each record. The id must be stable across uploads — rows with the same id update the existing record in Endear.

Timestamps

The updated_at column is optional, and in most cases we suggest leaving it blank. If you send an updated_at value, you must send an updated_at value in future uploads with a newer timestamp in order for data to be updated.

Deletes

Set deleted_at on a record to soft delete it. Soft-deleted records are removed from the active dataset but retained for historical reporting and analytics.

Money values

Monetary amounts are sent as decimal numbers, with the currency given in the currency_code column of the same row. A purchase line item's currency must match the currency of its purchase.

• For Product variants, if a price is unknown, simply list 0.01 as the placeholder price. This price is not used in purchase calculations.
• For Purchases, we perform a soft validation between subtotal, discount, shipping, tax, and total. The subtotal - discount + shipping + tax should equal or be similar to the total, but we understand there may be cases where it is slightly off.

Supported data types

Integration Imports support the data types below. Required types form the foundation Endear needs to operate. Recommended types aren't strictly necessary, but they unlock significant additional functionality and we strongly suggest importing them.