A few thoughts on Alan Jacobs' post on time management as an academic:
5) Many academics, in the humanities anyway, also over-comment on their students’ essays, and end up giving far more feedback than the students can absorb, even when they want to, which is not that often. If you write dozens of marginal comments and a page or more of summary comments, students will rarely be able to differentiate between the major issues and the minor ones. You need to make comments only about major things, and let the little ones go. In that way you’ll give your students feedback that they can actually use.
I found this to be particularly true with my freshmen. Too much advice, even well-meaning, gets lost. Better to focus on one, perhaps two, central tasks that need improvement, and leave everything else for later. It's sensible time management, as Jacobs identifies, and it's good pedagogy: the ability to write and control language comes in stages, and it does no good to let the person just beginning know they're doing everything wrong.
If more time is needed to work with students on writing (and more time is always needed), I found it best to collect repeated problems across the class and devote ten or fifteen minutes to, say, how to structure an introductory paragraph, or how to best use direct quotations. It gives the same benefit as concise comments--only gives one thing at a time to students--but increases the number of overall topics getting that kind of attention.
6) Also: I ask my students to give me, by email, a proposal two weeks before the essay is due. I tell them what I think is good about their idea and what they need to watch out for; more often than not I advise them to take only a part of their topic and focus on that. Then, a week later, I have them send me, again by email, a rough draft. Once more I comment briefly with encouragements, warnings, and indications of where they should invest their major energies. This process would be valuable to them even if I gave no comments at all, because it makes them think about their work well in advance of the due date, which gives them the chance to turn ideas over. By the time they turn in a final version, I don’t have to make many comments at all: those who put in the work will have improved significantly, and the others will already know what their problems are. I spend less time that I would have spent in writing extensive comments; I spread that labor out over a longer period, thus making it feel less onerous; and I get better results.
I've found that there is significant merit in opening the space for smaller interventions into student writing earlier on. This is why I have a very open email policy--far better to catch the problem in its infancy and redirect student attention to something better. It's also the motivation for building up end-of-semester writing projects in very small and discrete stages. For my upper level class, I had them do progressively longer assignments once a month in preparation for their term paper. This captures all the elements Jacobs discusses, and adds a few extra: it is obvious to me when a one-paragraph thesis is simply not going to work, in a way that is not obvious to the students themselves. But this allows for a five-minute intervention when there are still ten weeks to retool the paper, rather than looking at someone's draft and realizing it to be untenable a week before the final product is due. It also makes strict grading easier, since students who disregarded the advice I gave about how to best fit their topics within the guidelines of the paper and the space available to them have definitively earned the grades they receive.