Research & PolicyFree
Manuscript Formatting for Journal Submission
Send the manuscript and the target journal, get the reformatted file plus every requirement no edit to that file can satisfy.
River's manuscript formatting tool reads the target journal's author guidelines and its submission portal together, then turns them into a numbered list of discrete requirements instead of a wall of prose. It reformats the manuscript against the ones a file edit can close, and it flags the rest. What comes back is two artifacts: the reformatted manuscript, and a checklist saying which requirement each change answers, which guideline line imposed it, and the date that line was read.
Unlike a reference manager style or a journal's own Word template, it does not assume the job is the file. A style file fixes citations, and citations are the part authors report taking the least time of anything on the list. What is left is structure, declarations, and portal fields that no change to the manuscript will ever satisfy. Those get separated out and ordered by lead time, because they are what decides when you can press submit.
It is built for the corresponding author on a resubmission, the postdoc handed a rejected paper and a new target, and the coordinator running several submissions at once. Use it when a paper is bounced without review, when a co-author asks what is still outstanding, or when production sends a formatting list back after acceptance. Every run records the guideline line behind each requirement, so the journal after this one is a shorter job, the way a documented search strategy makes the second search shorter.
Why reformatting takes longer than the editing in it
The assumption is that reformatting is slow because there is a lot of it. It is slow because a small part of it cannot be done alone. Surveyed authors put readjusting references among the least time-consuming tasks on the list, and gave the largest share to publisher online requirements: contact details for every author, disclosures, the fields in the portal. Those are not edits. They are requests to other people, followed by waiting.
Take a paper bounced without review and pointed at a second journal. Its guidelines and portal produce 63 discrete requirements. 41 can be closed by editing the file, and 27 of those are already satisfied by the previous journal's format, leaving 14 real edits. 17 need a person or a document from outside the file. 5 cannot be satisfied by anything that exists today. So 51 close in one sitting and 12 do not.
Sort by lead time before editing anything. Those 12 fall into six waits, and the longest is a repository deposit quoted at ten working days, which sets the submission date by itself. Two ORCIDs and an ethics number account for three more. The 14 edits are not on that path at all. Start the requests on day one, do the edits while they run, and record the guideline line behind every requirement, so the journal after this one is a 19-item job rather than another 63 from scratch.
How it works
Paste the manuscript
Add the target journal and article type, and say whether the paper has been formatted for anyone before.
Requirements get extracted
Every line of the author guidelines and every field in the portal becomes one numbered requirement with its source.
The file gets reformatted
Only where the new guidelines actually differ, so nothing already compliant is touched or quietly broken.
Work the blocked list
Keep going in chat: chase the missing documents, cut to the word limit, decide what to do about a figure.
What you get
- Every requirement numbered, with the guideline line that imposed it and the date it was read
- The reformatted manuscript, changed only where the new guidelines actually differ from the old
- Requirements split by who can close them: a file edit, another person, or nobody yet
- Portal fields listed separately, since no change to the manuscript file ever satisfies one
- Blocked items named plainly, including the ones that are content decisions wearing a formatting label
- A lead-time order, so the requests that set your submission date start on day one
Common questions
Will it rewrite my science?
No. It changes structure, order, headings, citation style, captions and declarations. Where a requirement can only be met by cutting content, such as a word limit below your current length, it says so and stops rather than deciding for you. That is an authorship decision. Prose it does write, like a methods section, comes from what was actually run.
How long does this normally take by hand?
More than most people budget. A survey of 372 researchers across 41 countries put the median at 14 hours of formatting per manuscript, and 52 hours per person per year. Most of that is not writing. It is translating one set of arbitrary requirements into another, twice, because the median paper took two attempts to place.
Does it handle the submission portal fields?
It lists them and tells you which you cannot answer yet, which is the part that matters. It cannot log into a publisher's system and type on your behalf. What it gives you is every field in one place, with the ones needing a co-author, an office or an accession marked and ordered by how long each usually takes, the way a reporting calendar orders award deadlines.
My references are already in a reference manager. Does this help?
Yes, and for a reason worth knowing. Reference managers already solved the part authors find quickest, which leaves the slow part untouched: section order, structured abstract fields, figure resolution, declarations and portal metadata. The checklist covers those, and it will confirm your citation style rather than redoing it.
What if the journal's word limit is below my current length?
It is flagged as blocked with the exact gap, not silently trimmed. In the worked example a 4,000-word cap against 5,240 words meant 1,240 words had to go. Then you can keep working in chat on where they come from, which is a conversation about the argument rather than about formatting.
What happens when this journal rejects it too?
The trace is the point. Each requirement records the guideline line and the date it was read, so running the same extraction against the next journal produces a difference rather than a fresh start. In the worked example 44 of 63 were unchanged, making it a 19-item job. If it comes back with reviews instead, answer them here.
Which journals does it work for?
Any journal that publishes author guidelines, which is all of them. The extraction reads what that journal actually says rather than a stored profile, so a guideline changed last month is read as it stands now. Society journals with idiosyncratic requirements tend to produce the most useful checklists, because those are where surprises live.
Manuscript Formatting for Journal Submission
Fill in the form and your workspace opens with the work already underway.