After you’ve analyzed a model article from the journal you’re targeting, it’s time to write a conclusion to your own article. If that sounds crazy, bear with me.
Often, in my experience and in those I’ve talked with about article writing, it seems that people write in a fairly linear fashion — that is, they have some empirical material they want to work with, and so they plop it into a document and then set about writing an introduction for the article. Then they write the literature review, work through their material, and write a conclusion. The result of this is often that the conclusions that get written are very narrowly focused and the introductions are vapid — too many generalities in both. This is probably due to people not really having a sense of what the article is about when they draft their introduction, and so they tend not to say terribly much; and when it comes to the conclusion, you’ve spent so much time working through the material that you can’t see the forest for the trees.
Tackling a conclusion at the outset can give you a clear sense of your intended audience and the material you plan to cover in the article. So, in the most flat footed way, you might consider starting with a ‘In this article, I have argued that [insert argument here].’ And follow that with a ‘In order to make these claims, I have drawn upon my research on [insert data umbrella] to elucidate how [phenomenon X] operates in [situation A].’ Then give summaries of the evidentiary cases you’ve presented. Briefly overview them with an eye toward summarizing them and how they relate to your argument. Taken together, this should be about a paragraph or two — not much more than a page of double-spaced text.
What follows from that should be 2-3 paragraphs on the contribution that you’re making to the journal’s field of inquiry. I find it useful to think very practically about the scholars in the field who might be interested in the case. Consider a person or set of people as the focus for each of these paragraphs and answer for them why they should find the case generative to think through. And it can work well to scale up, so start with the most immediate, empirical body of literature and move up to more theoretical concerns.
For the first of these paragraphs, consider: who else is working on the same topic or in the same geographical area as you? What does your evidence help to show that extends the conversation about the topic or the field? This can be a good place to consider how your case is exceptional — that is, are people talking about what you’re talking about or not? If not, what does including your area of interest do to the subfield or study of a geographical place? If people are writing about what you’re writing about, how does your set of cases extend the conversation in new ways or point out previously overlooked factors? (This might take a couple of paragraphs if you’re contributing to both a subfield and a regional interest, but be sure to separate them — at least at this point — into discrete paragraphs.)
The succeeding paragraphs should extend your contributions to more abstract fields. Given your contributions to your subfield and geographic area, what are the more widespread debates you’re entering into? These debates shouldn’t be wholly divorced from interests in your subfield and geographic area, but should grow out of them, although they might be novel or surprising because your case is exceptional and points in new directions for future scholarship.
So, for example, if I’m writing about sleep for a medical anthropology journal, my first of these paragraphs might identify existing literature about sleep in anthropology and the social sciences, and what looking at this literature does from the perspective of medical anthropology. Then, I might write a paragraph that’s more focused on medical anthropology: what does focusing on sleep do to existing theories of medicalization, illness narratives, embodiment, etc.? And then, building out from there, I might write a paragraph about a more macro-level theoretical concept, say biopolitics, subjectivity, phenomenology, neoliberalism, etc. It’s possible to address many topics in your conclusion, but not so many that you try and address everything. Pick a few that are immediately relevant to the mission of the journal and focus your efforts there.
Generally, the last paragraphs of conclusions are the most far-reaching in their intent, often making suggestions about how your case and the insights it provides might affect the discipline more generally. If you’re following the suggestions so far, this is the culmination of your other contributions, and might be a distillation of them all and their potential, collective impact. So, extending my example, what does focusing on sleep in American society potentially do for anthropologists who aren’t interested in sleep or the U.S.? I always find this the hardest paragraph to write, in part because it depends on imagining a reader who is fairly alien to me. But if you can make your work relevant to that person — someone fundamentally disinterested in your case — then you have a pretty successful argument.
This should be a clarifying exercise. By the end of it, you should have a good sense of the content of the article (in terms of the evidentiary cases you’ll be using), as well as the sets of literature that you’ll be contributing to with your scholarship. In the past, when I’ve taken this approach — often with just the germ of an article in mind — it’s immensely helpful in identifying which journal I might send something to: based on the kinds of evidence, the imagined readers, and the overall contribution, I get a sense of whether or not I’m writing a long, highly detailed piece aimed at the discipline, or a short, targeted piece aimed at a particular subfield. (I know this contradicts my ‘analyze a model’ first step, but once you’ve analyzed a bunch of models from different journals, you can often start with a writing exercise like this one, and then move back to picking the right journal.)
Once you have your conclusion drafted, it’s time to move on to your literature review, which you should have already begun to think about by way of writing your conclusion. You can read about Step 3 here.
Matthew Wolf-Meyer 
I just wanted to say how incredibly helpful I find this series. I return to it often. I don’t always follow it, but it’s a very helpful touchstone for moving me out of writing paralysis. Thank you!
I just wanted to tell you how helpful I found this. I have read a lot (a LOT) about writing in general and writing journal articles specifically, but I have never read of this tactic of writing the conclusion first. I’m now doing this for the article that I’m currently revising, and I found it incredibly helpful in helping to hone my argument and make sure its properly tailored to the journal audience and the reviewers’ comments. Thank you!