Yogesh Singh
Most journal editors use Crossref for exactly one thing: assigning a DOI to a published article.
That works. But it means the vast majority of what Crossref can actually do is sitting unused – and some of the pieces being ignored are the ones that make a journal look complete to indexing bodies, funders, and databases.
Crossref currently accepts eleven distinct record types, plus components and relationship metadata that connect them together. Below is what each one is, who it’s for, and whether your journal should be registering it.
One thing to be clear about first, because it causes constant confusion: you never send Crossref your actual content. No PDFs, no full text, no files. You send a metadata record – title, authors, dates, affiliations, a URL – and Crossref links a persistent DOI to it. The content stays on your server. Crossref stores the description and the pointer.
The 11 content types Crossref accepts
1. Journals and articles
The one everybody knows. You can register at two levels: the journal title itself and each individual article. DOIs can also be assigned to volumes and issues, though most publishers don’t bother.
Registering the journal title is the step people skip. It gives your journal a DOI of its own and creates the container that all your article records live inside. If your title record is missing or wrong, every article underneath inherits the problem.
Should you register this? Yes – this is the baseline.
2. Components (supplementary material)
Components are the parts of an article that deserve their own identifier: figures, tables, supplementary files, appendices, videos, datasets attached to the paper.
Each component gets its own DOI, linked to the parent article. If someone wants to cite your Figure 3 specifically, they can.
Should you register this? Useful if you publish data-heavy or image-heavy research. Not urgent for a small journal, but it’s the cheapest way to look sophisticated.
3. Preprints and posted content
This covers anything made public but not formally published – preprints, eprints, working papers, and other posted material.
The valuable part isn’t the preprint DOI itself. It’s the relationship: when the article is finally published, you link the published version back to the preprint. Readers who land on the preprint get pointed to the version of record, and your citation counts stop being split across two objects.
Should you register this? Only if you run a preprint server or post accepted manuscripts ahead of print. But if you do post “online first” content, register it properly.
4. Peer reviews
You can register the reviews themselves – reviewer reports, editor decisions, author responses – as records attached to any work already in Crossref. Any number of them.
This is the single most underused type in the entire system, and it’s the one that most directly signals a serious journal. Publishing your peer review reports with DOIs is about as strong a transparency statement as a journal can make. Reviewers can be credited, or the record can stay anonymous – both are supported.
Should you register this? If you want to differentiate a young journal from the crowd, this is the highest-impact option on the list.
5. Books, chapters, and reference works
Full book-level records, chapter-level records, or both. Books can be registered as a monograph, a series, or a set.
Many society publishers and universities produce edited volumes, conference books, and textbooks that never get DOIs at all. Those books are then effectively invisible in citation systems, even though they’re cited in practice.
Should you register this? Yes, if your press or society publishes books. It costs little and the discoverability gain is large.
6. Conference proceedings
A record for the conference event itself, plus a record for each individual paper presented.
Every year, thousands of Indian and international conferences publish proceedings as an untracked PDF on a college website. Those papers are cited by attendees and then disappear. Registering them makes a conference permanently part of the scholarly record – and gives the organising institution something real to show.
Should you register this? Yes. This is the most commonly wasted opportunity in academic publishing.
7. Dissertations and theses
Individual dissertations and theses. Note the limitation: single works, not collections.
For a university, this is a straightforward way to make its research output citable and traceable, and to link student work to supervisors and departments.
Should you register this? Yes, if you’re a university or a repository.
8. Datasets
Database records and collections of data can be registered with Crossref.
There’s a genuine decision to make here, though. DataCite is the registration agency built specifically for research data, with metadata fields designed for it. Crossref is built around published literature. If you’re registering a lot of datasets, DataCite is usually the better home; if you’re registering a dataset closely tied to an article you’re publishing, Crossref keeps everything in one place.
Should you register this? Compare both before committing. Don’t default to Crossref just because you already have an account.
9. Reports and working papers
Published reports and working papers – typically the kind of output that has an ISSN and comes from a research institute, think tank, government body, or NGO rather than a journal.
Should you register this? Relevant if your organisation produces institutional reports. Ignore it if you only publish a journal.
10. Standards
Publications issued by standards bodies. Narrow category, but it exists so that standards can be cited persistently like any other work.
Should you register this? Only if you are a standards organisation.
11. Grants
Grants are registered by funders, not publishers, through Crossref’s Grant Linking System. They cover direct funding as well as other forms of support such as access to equipment and facilities.
This matters to journals indirectly but significantly: when funders register grants with DOIs and journals record funding information in article metadata, the two connect. Funders can trace what their money produced.
Should you register this? Not your job as a publisher – but do record funding information in your article metadata so the link works from your side.
12. Pending publications
A placeholder record with minimal metadata, used when a DOI needs to exist and be shared before the content is publicly available – embargoed work, or a DOI that must be printed in a document before publication day.
Should you register this? Occasionally useful. Good to know it exists.
Beyond record types: the metadata that actually matters
Choosing the right record type is the easy half. The half that determines whether your journal looks credible is what goes inside the record.
These fields are optional in the sense that Crossref will accept a deposit without them – and damaging in the sense that leaving them out is visible to everyone:
| Metadata | Why it matters |
|---|---|
| References | Deposit your reference lists and your articles become part of the citation graph. Skip it and nothing you cite links back to you. |
| Abstracts | Feeds discovery services and search engines directly. Enormous visibility gain, almost no effort. |
| ORCID iDs | Ties each author to a verified identity instead of a name string. Solves the “which R. Kumar is this?” problem permanently. |
| ROR IDs | Machine-readable institutional identifiers. Increasingly expected in place of free-text affiliations. |
| Funding information | Connects your articles to the funders who paid for the research. |
| Licence information | States the licence for each article in machine-readable form. Databases read this to confirm you are genuinely open access. |
| Full-text URLs | Required for Similarity Check, and used for text and data mining. |
| Relationships | Links versions, translations, corrections, datasets, preprints and supplements to each other. |
| Crossmark | Shows readers whether an article has been corrected, updated, or retracted. |
| Clinical trial numbers | Essential for medical journals. Links published results to the registered trial. |
Crossref publishes Participation Reports, which show publicly how complete any member’s metadata is. Anyone – including an indexing body assessing your application – can look up your journal and see exactly which fields you’re leaving empty. Most publishers have never checked their own.
What you cannot deposit
Worth stating plainly, because these come up constantly:
- The content itself. No PDFs or full text are stored. Only metadata and a link.
- Collections of theses. Individual dissertations only.
- Anything you don’t own or publish. You register content you are responsible for, under your own prefix.
- Content with no landing page. Every DOI must resolve to a page describing the work. A DOI pointing straight at a raw PDF is bad practice and breaks several downstream services.
And one obligation people forget: registering is not a one-time act. Keeping metadata accurate is a condition of membership. If your journal changes domain, every DOI must be updated to the new URLs. Updating existing records costs nothing – but broken DOIs will sink an indexing application faster than almost anything else.
What most journals should actually do first
If you run a journal and want a practical order of work:
- Register the journal title record, not just articles.
- Add references to every article deposit – this alone changes your citation visibility.
- Add abstracts, which cost nothing and improve discovery immediately.
- Add ORCID iDs and licence information to every record.
- Check your Participation Report to see what you’re missing.
- Then consider peer reviews and components as differentiators.
Steps 1 to 5 are configuration and habit, not budget. Most journals could complete them in a week and would look materially more professional to every database that reads Crossref metadata – which is nearly all of them.
Frequently asked questions
Can I deposit a book in Crossref if I’m a journal publisher? Yes. Your membership and prefix cover any content type you publish. You aren’t restricted to journals.
Do I pay separately for each content type? Yes – content registration fees are charged once per item registered, and rates vary by content type, with discounts available for back-year material. Updating a record you’ve already registered is free.
Should I use Crossref or DataCite for datasets? Compare both. DataCite is purpose-built for research data. Crossref makes more sense when a dataset is tightly bound to an article you’re publishing.
Can I register conference papers that were published years ago? Yes. Back-year content can be registered, often at reduced rates.
Does registering a DOI mean my journal is indexed? No – and this confusion causes real harm. A DOI makes your content identifiable and linkable. Indexing in DOAJ, Scopus or Web of Science is an entirely separate assessment of your journal’s quality and policies. Crossref registration supports those applications; it doesn’t replace them.
Getting this right
Most journals are registering perhaps 20% of what Crossref can do for them, and doing it with metadata that’s half empty. It isn’t negligence – nobody explains this to a new editor, and the documentation is written for engineers.
At DigitalBisht, we handle Crossref setup and DOI registration for journals and publishers: prefix setup, deposit configuration in OJS or custom platforms, metadata quality audits, back-year registration, and cleaning up records that were deposited wrongly the first time.
We’ve supported the launch of 100+ academic journals, and taken 30+ of them to Scopus indexing. If you’re not sure what your journal is currently depositing – or what it’s missing – send us your journal URL and we’ll check your Crossref records and tell you exactly where the gaps are.
