This is very much just the title question, post your workflow style. I have heard a few, some of them being strange. I use a common approach, but I will go into more detail.
1. Use a few spare brushes to create the bones of the map. If there is displaced terrain, I create a low detail terrain mesh.
2. Begin cutting and using more brushes to create rooms, making sure to using vertex manipulation to limit the number of faces seen.
3. A final pass through with brushes to create buttresses, supports, etc. Again, I am making sure to use vertex manipulation to conceal faces.
4. This is where I begin texturing. I may cut brushes and edit them to add textures where I may need them.
5. Props is next on my list. I usually prepare another vmf with all the props put down on it, so I can see what I may use at the same time, rather than one at a time in the 3D model viewer. I just switch between the main map and this prop map. I as place props, I also give each of them fade distances that are appropriate for their size and location.
6. Last in aesthetics is typically lighting & sound, along with other flares such as sprites, dust motes, fog, particles, decals, weather effects, and color correction.
7. Depending on who you are, you may have done this part much earlier, but I am one who likes to know what kind of world people will be playing in before they play the game itself. This is where I add the functionality and interactiveness, according to the game or gamemode.
8. Now that the map is actually in it’s final visual and gameplay state, I begin to optimize what has not been optimized in the process. Area portals go in, everything that can be func_detailed is done so, and surfaces with no shadows over them get large lightmap scales. Accordingly, surfaces with more intricate shadows over them get lower lightmap scales.
I’m interested in knowing how everyone goes about making their maps from start to finish, since we can all learn how to optimize our time spent in the Hammer editor.