How to Choose Reviewers to Suggest to a Journal

Topic overlap is the criterion everyone applies and it is the weaker half. A reviewer who works on your topic can tell an editor whether the question matters. A reviewer who has run your method can tell them whether the answer is right, and that is the judgement peer review exists to produce. When the two come apart, choose the method.

Ask what claim the paper stands or falls on, then ask who has tested a claim of that kind. That person is your first candidate.

Look at who your Methods section cites

the people whose technique, instrument or model you adopted can evaluate whether you adopted it correctly.

For interdisciplinary papers, cover the halves deliberately. Two reviewers who both understand the domain and neither of whom understands the statistics is a common failure.

Recency matters more than a long back catalogue

someone publishing on the method in the last three years knows the current literature and its known failure modes.

Being highly cited is not the same as being able to assess your work, and the most cited person in a field is usually the least likely to accept an invitation.

Independence is a factual test, not a matter of how you feel about someone. The question is whether an editor, or the reviewer themselves, would recognise a relationship that could bias the report. Work through the four checks below on every name before it goes on the list. The full typology is in the conflict of interest guide.

Co-authorship

search the candidate's name against every co-author on your paper, not only against your own. Three years is the window COPE uses and several publishers apply.

Institution

same university, same department, same research centre, same hospital group. Excluded almost everywhere, including for authors who have recently moved.

Supervision, in both directions and without a time limit in most policies: your doctoral supervisor, your postdoc mentor, your own former students.

Shared money

a current or recent joint grant, a shared consortium, a commercial interest in the outcome. This one is the easiest to overlook because it is often not visible in publication records.

A useful practical test

if you would have to explain the relationship in a sentence, the relationship is already too close.

The third criterion applies to the list, not to the individual, and it is the one most authors skip. Three reviewers from the same laboratory tradition will all miss the same thing, and an editor looking at three names from one country reads it as a network rather than a selection. Springer Nature recommends a mix of countries and institutions, and some Cambridge University Press titles ask that the suggested names span at least three countries.

Spread across institutions and countries

it is an explicit publisher expectation at several houses, and it also broadens the kinds of objection you will get.

Spread across career stages

mid-career researchers accept invitations more often and write more detailed reports than the most senior names in a field.

Cover the components of the paper

if the work has a clinical part and a computational part, do not send three clinicians.

Include at least one person likely to be sceptical of your framing. A report that engages seriously with an objection is worth more to the paper than three agreeable ones.

Do not build the list from a single journal's editorial board: it looks like an attempt to route the paper, and it narrows the perspectives you get.

Most authors already have the raw material and do not use it systematically. Four sources, in the order they usually pay off.

Your own reference list, filtered by conflict

the people you cite for the method rather than for the introduction are the closest match you will find.

The papers that cite the key work your study builds on, which surfaces groups working on the same problem right now rather than five years ago.

Recent programmes of the relevant conference, which show who is active and often reveal early-career researchers who are not yet highly cited.

A document-based search rather than a keyword search

our reviewer finder matches on the content of the manuscript you upload, which reaches groups who describe the same problem in different words. It suggests names and states the overlap; it does not check conflicts for you and it never contacts anyone.

A name that does not survive five minutes of checking will not survive the editorial office either. Run each candidate through this before the list is final.

Confirm they are still at the institution you are about to write down, using their departmental page rather than an old paper.

Take the email address from the institutional page or a recent corresponding-author line, and make sure the domain matches the affiliation you gave.

Check they have published in the last two or three years

a researcher who has left academia or retired will not review your paper.

Check they are not already in your acknowledgements, which is a relationship an editor can see and you have declared yourself.

Check whether they have publicly criticised your work. Disagreement in the literature is not a conflict of interest and is not grounds for exclusion, but it is worth knowing before you name someone.