← Integration Engineering
Integration Engineering

Data Mapping: “Customer” Means Something Different in Every System

7 Oct 2026·3 min read

Two systems can store the same person in very different ways. Moving data between them means deciding, field by field, what goes where. That is data mapping.

// System A
{ "id": "42", "full_name": "Sara Khan", "status": "Active", "phone": "" }

// System B
{ "id": 42, "first_name": "Sara", "last_name": "Khan", "is_active": true }
Example records, made up, not from a real API. Same person, four kinds of difference.

The four differences

  • Names: full_name on one side, first_name and last_name on the other. Where do you split "Muhammad Ali Khan"?
  • Types: the string "42" and the number 42 are not equal in code. A lookup finds nothing, and you create a duplicate.
  • Values: status is "Active" in one tool, 1 in another, true in a third. Where "Suspended" goes is a business question, so ask.
  • Empty: null, an empty string and a missing field can mean three different things. Turn "missing" into null and you may wipe a saved phone number.
Write the table before the codeFour columns: source field, target field, how it changes, and what happens when it is empty. Show it to someone who knows the business. It is the cheapest bug fix you will ever make.
Do this in 60 secondsPut one record from an API next to the same record in the system it feeds. Find one field for each of the four differences. That is the start of your mapping table.
Today’s AI angleAn AI agent calling several APIs has to do this mapping in its head, and it can guess wrong. Give it tools that already speak one clean shape, with the mapping done in code.
Data Mapping: “Customer” Means Something Different in Every System · Ali Huzaifa