A PID Policy Is a Workflow Requirement

A university can write "all outputs must have persistent identifiers" into a policy and still leave journals with the hard part. Which outputs? Which identifier agency? What metadata has to travel with the record? Who fixes the record when a funder, affiliation, author role, correction, repository link, or landing page changes? The policy sentence is short. The operating burden is not.

That gap is the useful warning in Crossref's July 20, 2026 position paper, "Persistent identifiers in research infrastructure policy: the need for a holistic approach": https://www.crossref.org/publications/pids-in-research-infrastructure-policy. Crossref argues that PID strategies should not be treated as a simple choice of identifier. What is really being chosen is infrastructure: open identifiers, rich linked metadata, interoperability, governance, services, sustainability, and long-term stewardship.

For journal leaders, this is not abstract policy work. It reaches into submission forms, production checks, DOI deposits, correction notices, repository relationships, funder reporting, author identity, and platform procurement. A PID policy that does not become workflow will eventually become a pile of exceptions.

The Weakest PID Policy Is A Noun

Many PID discussions still collapse into nouns: DOI, ORCID iD, ROR ID, grant ID, accession number. The noun matters, but it is not enough. A DOI with thin metadata is a locator, not a strong scholarly record. An ORCID iD entered as free text can be mistyped, orphaned, or disconnected from author permissions. A ROR ID captured at submission but dropped before publication is not an institutional signal. A grant identifier that never reaches Crossref metadata cannot do much work for funders or institutional reporting.

Crossref makes a sharper point: not all DOIs are equivalent because different DOI Registration Agencies provide different metadata, services, and governance. There is no central DOI metadata store that magically makes every DOI record equally discoverable or equally useful. If a policy says "use DOIs" but does not define the required metadata, relationship types, update obligations, and registration route, it leaves the most important implementation questions unanswered.

That is where journals feel the friction. The editor sees the article. The production team sees XML. The platform team sees required fields. The publisher sees membership obligations. The institution sees reporting. The funder sees compliance. The infrastructure sees a record that either connects across those contexts or does not.

Journal Managers Control More Than They Think

A journal may not set national PID strategy, but it controls many of the places where strategy either succeeds or fails. It decides whether ORCID iDs are authenticated or merely typed. It decides whether affiliations are normalized before acceptance or cleaned after publication. It decides whether funder data is structured, whether references are open, whether data and software links are retained, and whether corrections update the public page, the XML, the DOI record, and any downstream notices.

The important shift is to treat PID work as a chain rather than an endpoint. Submission captures identity and funding. Editorial review may add contributor roles, competing interests, data availability, ethics approvals, and reviewer governance signals. Production turns those facts into JATS, HTML, PDF, issue metadata, and registration packages. Hosting makes the record visible. Crossref, ORCID, ROR, DataCite, repositories, indexers, and institutional systems then read or reconcile what the journal preserved.

If one handoff breaks, the PID still exists, but its value drops. A DOI can resolve while the funding relationship is missing. An article can publish while the author affiliation is stale. A correction can be visible to readers while the metadata still describes the old state. The identifier did not fail. The workflow around it failed.

Non-Article Outputs Are Entering The Same Pipe

Crossref's Schema 5.5 release adds another practical prompt for journals and publishers. On July 9, Crossref announced Schema 5.5 with CRediT support and new record types for blogs and posters: https://www.crossref.org/blog/schema-55-now-available-adding-credit-new-record-types-for-blogs-and-posters-and-more/. That matters because scholarly communication is no longer cleanly divided between articles that get infrastructure and everything else that sits on a website.

Crossref has already explained why its own blog posts now carry DOI links and richer metadata, including authors, ORCID iDs, publication dates, archiving, versions, relationships, ROR IDs, and feed-based registration through Rogue Scholar: https://www.crossref.org/blog/scholarly-blogs-and-their-place-in-the-research-nexus/. ORCID also added support for "blog post" as a work type earlier in 2026 to align more closely with COAR resource type vocabulary: https://info.orcid.org/new-work-types/.

This does not mean every society newsletter item needs a DOI tomorrow morning. It does mean that journals should decide which non-article materials belong in the scholarly record before authors, editors, or communications teams start making ad hoc choices. Editorials, policy statements, methods notes, posters, data notes, reviewer acknowledgements, conference outputs, protocols, and research commentary can all carry scholarly value. The question is whether the journal has a consistent way to identify, describe, preserve, and update them.

Persistence Means Someone Can Repair The Record

Persistence is often heard as a technical promise: the link will keep working. In practice, persistence is a maintenance promise. Someone has credentials. Someone knows which landing page is authoritative. Someone can update metadata after a platform migration. Someone can correct an affiliation, add a funder relationship, mark a retraction, or preserve access when a journal transfers publisher.

That is why the Crossref paper's emphasis on governance and sustainability matters for ordinary journal operations. A persistent identifier strategy fails quietly when staff turnover leaves no owner, when vendor contracts do not specify metadata portability, when legacy article pages redirect badly, when corrections are treated as page edits rather than registration updates, or when a publisher cannot explain which system is authoritative for article relationships.

The repair path should be designed before it is needed. If an author changes affiliation after acceptance but before publication, which record wins? If a data repository DOI is added after review, who validates it? If an article is withdrawn after Online First publication, who updates the DOI record and public status? If a journal moves to a new domain, how quickly can landing pages, sitemap records, canonical URLs, and DOI targets be checked? These are workflow questions, not metadata trivia.

Where Publicator Fits

For publishers that need this discipline across more than one title, Publicator can keep PID work attached to the publishing workflow instead of scattering it across spreadsheets and vendor tickets. AI-assisted submission checks can flag missing identifiers early; reviewer matching and governance can stay linked to audit trails; DOI and Crossref-ready metadata can move with JATS, PDF, and HTML production; journal hosting, role-scoped access, SSO, data residency controls, analytics, integrations, and multi-journal management can help teams see whether stewardship is consistent across the portfolio.

A 30-Day PID Walkthrough

• Pick five recently accepted articles and trace ORCID iDs, affiliations, funders, grants, data links, software links, references, contributor roles, and corrections from submission to the public page and DOI metadata.
• List every output type the journal publishes besides research articles, then decide which ones should be registered, preserved, or explicitly left outside PID workflows.
• Check whether DOI, ORCID, ROR, funder, repository, and related-object fields are structured data, authenticated where possible, and retained through production exports.
• Name the person or role that can update each PID-facing record after publication, including weekends, staff transitions, publisher transfers, and platform migrations.
• Ask vendors for the evidence, not the promise: sample deposits, update logs, schema support, Crossref reports, XML validation output, and documented correction paths.

The walkthrough will expose two kinds of problems. Some are missing fields. Those are annoying but usually fixable. The deeper problems are missing ownership and missing handoffs. If nobody knows who updates a relationship after publication, the system is not persistent in the way policy writers usually mean.

Practical Takeaway For Journal Leaders

The practical takeaway is to translate PID policy into a workflow map. For each identifier or relationship your journal relies on, write down where it is captured, who verifies it, where it appears, which export carries it, how it is updated, and who owns repair after publication. Do this for articles first, then repeat it for the non-article outputs that now deserve a clearer place in the record.

A journal does not earn infrastructure value by sprinkling identifiers onto finished pages. It earns that value by keeping the metadata, relationships, and correction paths alive after the identifier is minted. That is the real work behind a PID policy, and it belongs in the journal operation before the next mandate, migration, or audit makes the gap visible.