Challenge 2: Mismatched Column Names Without Match Fields — Solution Walkthrough What happens without using Match Fields first: Word's merge field list is built directly from whatever the ACTUAL column headers in the connected Excel source are named — CustName, AmtDue, Due_Date — not from some fixed, universal set of names it expects. If merge fields are inserted using THOSE actual source names, the merge would technically work correctly, since the fields genuinely exist with those names. The real risk only appears if the TEMPLATE was written assuming a different set of names (Name, Amount, DueDate) — for example, the template was copied from a previous merge project that used those names, or those exact field names were typed in expecting Word to somehow know they mean the same thing as CustName/AmtDue/Due_Date. In that case, «Name», «Amount», and «DueDate» would not correspond to any actual column in THIS source at all — Word either shows an error identifying the unrecognised field name, or in some cases silently leaves that portion of the merge blank, since there is no genuine automatic connection between differently-worded field names without an explicit mapping step. The correct fix: Use Mailings > Match Fields to explicitly map Word's own standard field roles (First Name, Last Name, Address, etc. — the built-in role list Match Fields offers) onto whichever of the source's actual columns (CustName, AmtDue, Due_Date) correspond to each one. Once mapped, inserting a merge field via its Word-recognised role name correctly pulls from the right source column regardless of what it's actually called in the spreadsheet. WHY THIS WORKS AS AN ANSWER ------------------------------ This is exactly the exact-match discipline this chapter compares directly to XLOOKUP: Word cannot infer that "AmtDue" and "Amount" refer to the same underlying concept on its own — an explicit mapping step (Match Fields) is required to bridge that gap, the same way an XLOOKUP's lookup value and the column it searches must match exactly, with no fuzzy inference happening automatically on your behalf.