A field name is not a complete rule
A useful rule records the source value, its destination, any transformation and the check that will prove it worked. Add what should happen when the source value is missing or cannot be represented. That last column is where many migration decisions become visible.
For example, “created date → date” is ambiguous. A record may have a creation date, first publication date and last-modified date. Choose the one the destination should display, record the source timezone and test the resulting page.
The destination must also support the field. In WordPress, custom content types and taxonomies need the appropriate REST configuration when that access path is used. The WordPress REST handbook describes that configuration; it does not mean that every plugin field will transfer automatically.
Follow article 42 through a small mapping demonstration
We used three synthetic source articles, one known author and two categories both named “News”, under different parents. A deliberately faulty first run reused source author IDs, chose the first category with that name and substituted the import date. The corrected run resolved the intended destination identities and preserved the publication timestamp.
| Source | Faulty result | Corrected result |
|---|---|---|
| Author 7: Demo editor | ID 7; wrong destination account | ID 107; Demo editor byline |
| Category 12: Library / News | ID 211; Company / News | ID 212; Library / News |
| Published 6 April 2020 at 09:30 +02:00 | 13 September 2026, the configured import date | 6 April 2020 at 07:30 UTC, the same instant as the source |
The third article, ID 44, refers to unknown author 999. The corrected run holds it for review and imports the other two records. Automatically assigning it to an administrator would hide the unresolved authorship.
Download the three source records and the observed results. This is a local teaching example with a simple destination model, not a tested WordPress, Drupal or Strapi connector. Its result shows why identity and date rules need checking against the rendered page.
For your project, use the mapping worksheet to record the actual field names, acceptance rules and exception decisions. Include media and alt text checks too; the small executed demonstration does not test files.
Load dependencies before the records that use them
Authors and taxonomy terms often need destination identifiers before an article can refer to them. Parent terms need to exist before their children can be assigned. Files may need uploading before the body can point to the new address.
Keep an explicit old-ID to new-ID map. Matching only by name can merge unrelated records; copying an old numeric identifier into a destination relationship can point to the wrong record entirely.
Some relationships form cycles. Two articles may refer to each other, for example. A staged approach can create the records first and connect them afterward, provided the destination and the agreed migration path allow it. Test that approach on a small sample.
Review formatting as well as stored values
A body field may contain HTML, shortcodes, rich-text data or nested components. Decide which parts the destination can render. Stripping every unfamiliar value may make an import pass while removing useful content.
For WordPress to Strapi, agree whether an embedded block becomes body content, a component or separate development work. For Drupal to WordPress, review compound fields and Paragraphs before selecting WordPress fields or blocks.
Use the trial to inspect what the reader sees and what the editor can maintain. A value stored in an unused field has not solved the editorial requirement.
Make exceptions a decision, not a silent fallback
Choose a response for missing authors, invalid dates, duplicate slugs and inaccessible files. Depending on the record, that might be an approved fallback, a blocked import or a documented exclusion. Record who can accept the exception.
Keep private content private while resolving missing access information. If the destination cannot represent the required restriction, stop that part of the transfer and agree a suitable approach.
Once the rules pass the trial, version the worksheet and use the same revision for the final run. Late changes need another check against the affected records. The migration methodology explains where that review belongs.
Use the worksheet to make the estimate more useful
An archive with many consistent articles can need less custom work than a small collection of heavily structured records. Showing the field and exception rules helps explain that difference before the project is scheduled.
Send representative content types and the proposed destination model with your enquiry. The cost guide and timeline guide describe the other dependencies. At handover, compare the result with the same acceptance rules in the validation report.