How to revise:

As an editor and author I have seen many revised papers return to journals. Given effort, most go well (ie step toward acceptance). Some go pear-shaped. I’ve slowly improved and have an approach known by my group as the ‘Breakspear method”. Here is its essence
1/ Aim for 1 round of revisions. Make the 1st response a big one. Be prepared to do as much work on the revisions as you did for the paper. It might be an overshoot, but it’s way better than going back again, which gets messier each time, or even worse, a “revision rejected”
2/ Approach the revision as a way of improving the paper, not as a way of placating the editors and reviewers. Despite it's caveats, IMO constructive peer review followed by careful revisions almost always makes the paper better: More accurate, clearer and better contextualized.
3/ Share the decision letter with your co-authors and read all the responses carefully. Digest over a few days, reading a few times. Then schedule a meeting with the key co-authors to come up with a strategy for the entire response. Do not get stuck into revisions one by one.
4/ Decide on & prioritize the new work to be done. *The* key to success on the next round is actually the new work. Editors + reviewers get happy at its sight. Be prepared to put new material into the main text not just in Supp Inf. Do the work as thoroughly as the original work.
5/ Treat the reviews as the most careful technical, independent reading you will get. If you think the reviews are off target, don't get mad. Aim to make something good out of every point. If the reviewer is misguided take this as a sign that the text wasn’t sufficiently clear
6/ Focus on those critiques common to > one reviewer and try to integrate your responses to these. Cross-reference between reviewers in the response letter (“see also R2.1”) but also make sure responses to each reviewer are self-complete (don’t let R3 think you ran out of steam)
7/ Pick your push-back carefully. Often there is some work that can't be done, is truly out of scope or is an error by the reviewer. The more overall effort, the more leverage to push-back on a few points. Politely state your actual response and the reason you did not go further
8/ The response letter should be self-complete. Add a brief summary of the main changes at the top of the document and put this in the cover letter. Use distinct text for each of the reviewers’ comments (R1.1, R1.2 etc), the *brief* responses, and then the new text + results
9/ Let the new work speak for itself and keep the rejoinder as brief as possible. There is no use in writing a long response to the reviewer with nothing in the revised ms. No reviewer ever writes a critique b/c they want a personal response from the authors!
10/ The best type of response is something like;

R2.3: The authors made an error etc

Response: The reviewer is correct. We have performed additional analyses and revised the manuscript accordingly (pX),

Edits: [new text in revised ms together + any new results and figures]
11/ I tend to quote most of the new text in the response letter and not just point the reviewer to the edited page/line. The goal is to allow the reviewer to read the response letter in full, & have a really good (hopefully positive) impression before they read the revised ms.
12/ Thank the editors & reviewers once at the top of the response letter, not with every response. Reviewers don’t write a critique as a favour but rather to improve the study. The best gratitude you can show is by taking the reviews seriously and revising the paper accordingly.
13/ Always use impartial language. NEVER get emotive, no matter how mad you feel. Take a deep breath, 'feel the good karma' and make an objective response. Make changes “to improve/correct the manuscript” not to “placate the reviewer” etc
14/ Make sure the paper keeps its narrative and logic. Don’t drop new material in ad hoc just to placate the reviewers. Edit it in seamlessly & then read the paper from start to end. Readers will have limited or no knowledge of the various exchanges of the review process.
15/ Make sure you resubmit a manuscript with all the edits clearly marked (say in blue or red text) and a cleaned-up version (depending on journal requirements) + a cover letter with the main revisions (not the original one) + the response letter
Good luck! I hope this helps with your next revisions, and also helps keep your experience of science enjoyable, constructive and educational, as it is (most of the time!) for me :-)
Make sure to post any questions and I will do my best to respond :-)
You can follow @DrBreaky.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: