Below is my itemisation of the “Marvel Method” of making a comic-book, as best as I can make it out from reading various interviews from Stan Lee / Jack Kirby glory days.
The method began at Atlas and would later change and morph at Marvel in the 1970s, but here’s what it was meant to be originally in the early days of Marvel…
1) Stan Lee wrote (or told face-to-face, sometimes with physical play-acting) a quick loose plot treatment for the comic’s next issue, ranging in length from a simple paragraph to a page or two. Just a plot with beginning – middle – end, and some general indications of where each scene might be set and what the character motivations/reactions might be at specific points. There were almost no detailed descriptions for the artist to follow, unless Lee wanted something very specific or innovative in a particular scene. The ending might also be made quite specific, in terms of exactly how things should end up in the final scenes.
2) The artist would break this story down into pages of framed art, pace the story across the number of pages available (sans splash page). The character designs and relationships/motivations had of course already been set up by previous issues of the title, so the artist could work pretty much ‘on automatic’ in that respect. Then he laid out the story across each page, with rough pencils. He might have to micro-plot specific scenes that the wider plot seemed to require, and/or bring a secondary character into the scene if that was needed. He added visual emphases and visual “cliffhanger” moments.
Obviously here the “Marvel Method” assumed a top-flight action artist like Kirby etc, as the method is not going to work with talent that has lesser visualisation and pacing skills. Or with comics that talk-talk-talk rather than show. Or with comics that need elaborate multi-issue pre-plotting and fiddly sub-plots.
3) The artist of course knows to leave space on the penciled pages for caption boxes or dialogue balloons, but doesn’t indicate exactly where they would be. He likely adds margin notes on the top of the page, to: i) explain the action to the writer, if the pencils are very rough or he’s invented a filler scene; or ii) repeat the section of the writer’s plot he’s covering, to jog the writer’s memory and/or for future reference later down the production line. That would be especially important if the plot had been delivered verbally.
4) The artist then added the splash (opening) full-page to summarise the story, meaning the story as they had imagined it on the page.
5) Stan Lee then came back into the process. He approved the penciled art pages for conforming to his basic plot, picked up errors and misinterpretations (most likely via notes to the inker, such as “the people in this cafe need to be more seedy-looking, please”).
7) Stan Lee penciled in the placing and size of the caption boxes, and suitably sized balloons, so as to meld these well with the artwork. Lee also said in one interview that he invented and penciled in all the sound-effect words (THHWONK! etc) at that point. This means the writer must also have an excellent sense of graphic design and spacing in page layout.
8) Once that was done he had a good idea of how wordy each page could be. He then wrote the copy for the caption boxes and dialogue/thought balloons, so as to fit with what was in the panels — the facial expressions, the details shown in the settings, any new elements the artist added etc. His captions would also help fill in plot aspects that were not being conveyed by the visuals.
Stan Lee: “And I found, as I was doing it, it made it much more enjoyable. Because I wasn’t looking at blank paper in a typewriter, but I was writing copy [dialogue, captions] for people, for drawings that I was looking at, with expressions and actions. I felt carried away.” [He would also ‘act out loud’ his ideas for lines, trying out how they sounded, in the same way that Robert E. Howard had done in the 1930s].
9) Small gaps left over in the layout might be filled with occasional footnotes to the reader, which referred them back to events in previous issues or in other titles. This helped to cross-sell titles to readers, and also helped the creatives keep track of continuity.
10) Did Lee also pencil in the lettering of his copy-writing? I guess it would have saved him typing time, and it seems he did — as in one interview he admitted: “I write the captions and dialogue [aka ‘the copy’], usually right on the original artwork”. This would also have helped with exactly fitting the copy to the space available on the page, making the finished page look even more elegantly arranged. Which means the writer also has to have lettering and copy-fitting skills, albeit not in ink.
The penciled and laid-out and written pages were presumably then photographed as a backup, and then went off to the inker of the artwork. Then to a letterer and colourist, under the supervision of an editor.
That, as best I can re-construct it, is how it was done.