Build a Media Source Inventory Before Combining Libraries

A source inventory describes each authorised source as a system, exposing overlap, metadata differences, access risk, and migration dependencies before libraries are combined.

In short: Create one inventory row per authorised source before combining any library views. Record ownership or authorisation, access status, content profile, approximate scale, metadata quality, language coverage, known overlap, dependency risk, and the next safe action. Never put passwords, tokens, or private access details in the inventory.

A source inventory answers “what systems are involved?” before item-level cleanup answers “what should happen to this title?” That order prevents accidental duplication, source removal, and unsupported assumptions about which copy is authoritative.

Inventory sources, not credentials

Use a neutral source label such as “Household source A.” Record:

• owner or authorised user;
• authorisation confirmed: yes, no, or needs review;
• connection status;
• media types;
• approximate size or broad scale;
• languages represented;
• metadata strengths and gaps;
• update frequency;
• backup or recovery owner;
• known end date or change risk;
• safe next action.

Store credentials only through the source's approved secure process. The inventory may point to a recovery owner, but it must not contain a password, one-time code, token, or secret address.

Use collection-level descriptions first

NARA defines an inventory as a descriptive listing of record groups or systems with location and other relevant data, rather than necessarily every document. Its rules apply to government records, not household media, but this distinction is useful: begin with systems and groups, then inspect items only where a decision requires them.

The Library of Congress also treats inventory and custody as distinct management concerns for digital collections. In a household plan, identify who is responsible for each source and who may approve changes.

Sample each source consistently

Choose five known items from every source:

1. a common title likely to overlap;
2. an item with strong metadata;
3. an item with missing or weak metadata;
4. a multilingual item when relevant;
5. an item that changed recently.

Record what the source exposes: title, type, date, language, version, availability, and category. Do not infer a field that is absent.

This sample is not a statistical survey. It is a practical way to expose different source conventions before migration.

Map overlap without deleting duplicates

Create three overlap states:

• Confirmed: same work and comparable version are clearly present in two sources.
• Possible: titles appear related but version, language, date, or edit differs.
• Distinct: evidence shows the items serve different needs.

Do not deduplicate from title alone. Two items can share a title but differ by cut, language, quality, rights, or availability. Preserve the source relationship for later review.

The library migration plan explains how to move or combine views without turning overlap into new clutter.

Identify source dependencies

For each source, ask:

• Does the collection depend on an account that may expire?
• Is access tied to one household member?
• Can the source export useful metadata?
• Are categories source-specific?
• Is a device or application required to read it?
• What happens if the source changes names, structure, or availability?

Record unknowns explicitly. The source change resilience guide turns them into fallback actions.

Original evidence: source inventory register

Add a separate sample-items sheet using non-sensitive identifiers. Compare the register with the collection scope and remove any source that is not authorised or not part of the current phase.

Common mistakes and limitations

• Storing credentials in the inventory.
• Listing every title before understanding the sources.
• Assuming the largest source is authoritative.
• Calling two matching titles exact duplicates without version evidence.
• Ignoring who owns or may change a source.
• Combining categories whose meanings differ.
• Removing a source before documenting dependencies.

Norva organises a compatible source the user owns or is legally authorised to use. It does not create source rights or guarantee that two sources expose equivalent metadata.

Frequently asked questions

How accurate must the item count be?

Use a broad or source-reported figure when exact counts are expensive. The inventory is for planning; mark the method and date so the figure is not mistaken for a permanent fact.

Should inaccessible sources stay in the inventory?

Yes, if they are relevant, but mark access and authorisation as unresolved and do not combine them into the active plan until verified.

When should I move to item-level work?

After source ownership, scope, and overlap risks are understood. Item-level review should answer a specific decision, not become automatic busywork.

Your next step

See how Norva connects a source

Sources

• National Archives: Records inventory introduction
• National Archives: Knowing your records
• Library of Congress: Inventory and custody
• How Norva works