Match every output to its receiving role.
ENG-024_0007.tifIdentify the actual target environment
Start with the product and version, the named destination such as a site, library, vault, object store, file cabinet, or folder, and the customer administrator responsible for access. Identify content types, templates, document classes, object properties, required fields, controlled values, folder rules, file restrictions, and user or ownership mapping needs where applicable.
A platform name does not define the import path. SharePoint, for example, has documented migration tooling with specified source and destination requirements. OnBase uses configured scan queues, Document Types, and Keyword Types. Other environments may use an approved API, an import-index file, a watched process, a migration tool, or a customer-run procedure.
Design files, metadata, and manifest together
The content package should identify each file and its role: captured image, access PDF, validated PDF/A where specified, OCR output, or another requested derivative. The metadata map should define target fields and formats. A manifest or sidecar should join a stable record ID with the relative file path, MIME type, source reference, hierarchy, OCR status, and delivery state.
Fictional example — no PII: record `REC-000231` can link `images/REC-000231.tif`, `pdf/REC-000231.pdf`, a title, record date, source roll and frame range, and an OCR-status flag. The destination may transform this neutral record into a repository-specific import structure after its rules are confirmed.
| Planning item | Question to settle |
|---|---|
| Content files | Which source and derivative formats are needed? |
| Record ID | What stable value joins files, OCR, and metadata? |
| Field map | Which source values map to which target fields and types? |
| Import path | Which approved tool, API, index file, or process will be used? |
| Acceptance | How are required fields, file associations, counts, and exceptions checked? |
Use a representative test batch
A test batch exercises real field values, controlled lists, dates, file relationships, and exception cases in the selected workflow. It can confirm whether a file name, path, multi-page TIFF, PDF, text layer, or metadata format needs adjustment before a larger release. The test does not imply that a generic package automatically fits every tenant or repository configuration.
Document the test scope, results, owner of each action, and the approved changes. If the destination has an administrator-managed API or migration service, access and configuration remain the responsibility of the appropriate customer team. A release plan should state who performs the load and how exceptions are reconciled.
Define acceptance and reconciliation
The final delivery can include file counts, record counts, checksums when requested, manifest totals, exception lists, and a data dictionary. Reconciliation checks that expected records are present and that any intentionally excluded, unreadable, duplicate, or exception-routed material is explicitly accounted for. It is a project-defined delivery control, not a claim of automatic import success.
National Microfilm can prepare files and metadata for a customer-approved destination. Discuss the target platform, version, field requirements, formats, access owner, and test criteria so delivery is aligned to the real workflow rather than an assumed universal connector.
Test a record-ID manifest and field map before wider delivery
A useful test batch keeps each content file connected to a stable record ID, source-media reference, selected derivative, OCR status and exception record. The data dictionary defines required fields, data types, date formats, controlled values and transformations. Compare the files and metadata against the actual receiving workflow rather than treating a CSV filename as proof of import readiness.
SharePoint library fields, Laserfiche templates, OnBase Document Types and Keyword Types, DocuWare cabinets, M-Files properties, OpenText metadata and FileNet classes can require different mappings. The receiving administrator confirms the approved method, permissions and acceptance checks. A neutral manifest or sidecar can support that work, but it is not a claim of a native connector, automatic load or universal searchable output.
- Check stable IDs, source references, file associations and required metadata together.
- Document derivative roles, OCR limitations and exceptions in the sample package.
- Have the destination administrator confirm the field map and test results before the agreed release.
Common questions
Can one metadata CSV be imported into every system?
No. A neutral CSV is useful for mapping and reconciliation, but repository-specific workflows often require different field, path, or configuration rules.
Who should define required fields and controlled values?
The customer’s repository and records stakeholders should approve them, with the data dictionary documenting the final map.
What does a test batch prove?
It checks the selected package against the agreed workflow and identified acceptance criteria; it does not prove universal compatibility.
Sources & further reading
These references explain technical formats, standards or destination requirements. They do not imply endorsement of National Microfilm.