Choosing a publishing system for an academic journal: an editor's guide

When choosing a publishing management system for an academic journal, five areas matter most: the path a manuscript follows from submission to publication, how editor and reviewer roles are separated, how issues are produced and published, whether the archive and metadata meet the quality indexes expect, and whether the journal can publish on its own domain. The right system takes technical chores off the editor's desk and leaves time for scientific decisions. This guide goes through the questions to ask in each area.

Where should you start when choosing a publishing system?

Start with the journal's own process, not with software. How many manuscripts arrive, how many editors work on them, whether review is single-blind or double-blind, how many issues appear a year, in which languages the journal publishes and which indexes it aims for. The answers make expectations concrete and keep attention away from impressive features that do nothing for you.

Journals often find themselves choosing between a platform shared with other journals and software whose installation, updates and security fall on the journal's own team. The first is easy but offers limited control over brand, domain and workflow; the second is flexible but technically demanding. A sound evaluation weighs today's needs together with the maintenance burden five years from now.

• Does the system clearly define the journal's manuscript states and decision options?
• Can editor, section editor, reviewer and author roles be separated, and can one person hold more than one role?
• Is blind review left to the editor's vigilance, or enforced by the system itself?
• Can it produce the metadata and export formats indexes expect for every article?
• Can back issues be imported in a controlled way, and rolled back if needed?
• Can the journal publish on its own domain, under its own identity?

It is also worth asking at the outset who owns the data. Articles, review reports, author records and publication history are the journal's scholarly capital; even if the system changes, this data must be exportable in full and in standard formats. Keeping each journal's data separate is an important safeguard for both security and portability.

A good publishing system frees the editor from technical chores and always leaves the decision to the editor.

What should the manuscript workflow look like?

The workflow defines which states a submission passes through and who can move it from one state to the next. A typical flow runs through submitted, in review, revision, accepted and published, with rejection as a branch that can occur at any stage. When these states are clearly defined in the system, no manuscript gets lost in between and every transition is tied to a role.

The information collected at submission is the source of the metadata the article will carry later. Each author's name, affiliation, email and ORCID iD should be captured in separate fields, and the abstract and keywords in each of the journal's publishing languages. Collected correctly at the start, none of it has to be retyped at publication.

There should be an initial screening step between submission and reviewer assignment. At this point the editor checks whether the manuscript fits the journal's scope, follows the author guidelines, has an acceptable similarity report and has been properly anonymised for double-blind review. Returning out-of-scope manuscripts at this stage, with reasons, saves both authors' and reviewers' time.

Revision rounds should continue on the same manuscript record, with each round's files, review reports and the author's response letter visible in order. Sending decision letters by email from within the system, and storing them there, keeps all correspondence with the author in one place.

How should editor and reviewer roles be structured?

A journal has more than editors and reviewers. The editor-in-chief, section editors, the editorial board, language editors, layout editors and proofreaders each need different screens and different permissions. The system should keep these roles distinct while allowing one person to hold several; in small journals it is normal for an editor to be the language editor too.

Scope matters as much as role. A section editor who sees only the manuscripts in their own section has a manageable workload, and confidentiality is preserved. Likewise, a reviewer should see only the manuscript assigned to them, with no access to other reviewers' comments or the details of the decision.

In journals that use double-blind review, author identity must not reach the reviewer through any channel: not on screen, not in a file name and not in automatic emails. Having the system enforce this, rather than relying on the editor's attention, is an important safeguard for publication ethics.

The reviewer pool should be treated as part of the role structure too. When reviewers' areas of expertise, keywords and completed and ongoing reviews are recorded, editors find the right person faster and avoid overloading the same reviewers. Tracking invitations and their acceptance or refusal in the system also makes late reviews visible early.

How are issues and production managed?

An accepted article goes through production before publication: language editing, layout and proofreading. Assigning each step to a named person and being able to track its status makes it easy to see what each article is waiting for as the issue closes.

Issues are defined by volume, number and year. Early view, which lets an article appear before its issue is finalised, shortens the gap between acceptance and publication and matters a great deal to authors. When an issue closes, a table of contents, a full-issue PDF and a cover image are needed; generating these from article data reduces manual errors.

Every published article needs a readable page: abstract, authors, affiliations, keywords, PDF, citation details and DOI. Letting readers copy the citation in different styles and download it in formats their reference managers accept makes the article easier to use and to cite.

Why are archive and metadata decisive?

A journal's visibility depends on how well its articles are described. Search engines and academic search services read the structured metadata on article pages, and indexing bodies expect files prepared in specific formats. Missing or inconsistent metadata can leave even a strong article undiscovered.

• Every article page should carry the citation_ meta tags that Google Scholar reads.
• schema.org markup, a sitemap and an RSS feed help search engines find new articles quickly.
• The XML Crossref expects for DOI registration should be generated by the system from article data.
• Exports in DOAJ, PubMed and JATS formats remove the manual preparation burden from index applications.

The archive needs the same quality. Back issues are the biggest worry for journals moving to a new system. Being able to import them from Crossref or JATS data, preview the import beforehand and roll it back in one step if something goes wrong greatly reduces the risk of migration.

Metadata quality is about more than format. Author names must be spelled the same way in every issue, institution names used consistently, abstracts and keywords complete in each publishing language, and article pages must match the PDF exactly. Publishing editorial turnaround times on the journal site, such as the average time from submission to decision and from acceptance to publication, is also a form of transparency that reassures authors.

Should the journal publish on its own domain?

A journal's own domain is part of its identity. Authors, reviewers and indexing bodies know the journal by that address; when it changes, old links break and hard-won visibility can be lost. A domain owned by the journal keeps the address stable even if the system or provider changes.

Technically this is usually done with a CNAME record: the journal's domain is pointed to the publishing system in the domain management panel, and the system binds it to the journal once the record is verified. It should also be written down who in the institution owns the domain, who tracks its renewal and who holds DNS access.

For journals that use DOIs, a domain change also means updating DOI records to the new address. Choosing a permanent address, and setting up permanent redirects from old addresses when a move is unavoidable, is essential for the integrity of the archive.

Frequently asked questions

Does a small journal need a publishing management system?

A small journal publishing a few issues a year still has to keep its workflow, reviewer confidentiality and metadata in order. In a small team, running all this by email means the editor spends most of their time on correspondence, so a system is if anything more valuable for small journals.

Do old articles' DOIs change when moving to a new system?

No. A DOI is a persistent identifier and does not change. What changes is the address it resolves to; after migration, the addresses in the Crossref records need to be updated to the new article pages.

Should the journal website and the editorial desk share an address?

Having both under the same domain is convenient for authors and reviewers, who see their own screen according to their role once signed in. What matters is that editorial screens are open only to authorised roles and do not follow the same access rules as the public pages.

How much does the journal website's design matter?

Design shapes the first impression of how serious a journal is, but readability and accessibility matter more. Article pages that read well on mobile, PDFs that open in the browser, a searchable archive and an easy-to-find editorial board and policy pages count for more than striking visuals.

How much does changing the publishing system disrupt a journal?

In a well-planned migration, the visible change for readers and authors is limited. Importing back issues with a preview, setting up permanent redirects from old addresses, updating DOI targets and deciding in advance where ongoing reviews will be completed make it possible to switch without disrupting the publishing schedule.

Before you decide

When evaluating a publishing system, rather than relying on demo screens, take one of your own manuscripts through the whole process: submission, reviewer assignment, revision, acceptance, production, placement in an issue and publication. That trial shows how well the system fits your process better than several meetings would.

A well-chosen system carries a journal's identity, archive and editorial quality for years. At Bir Web Ajans we develop Nova JournalPort, which gives every journal its own site, its own database, its own domain and a complete editorial desk, and we plan journals' migrations on that foundation.