The plan shows you what to build. It does not show you when.
That's the part that gets people. They've got the cut list. They've got the dimensions. They've got a nice exploded view with callouts. And they start cutting. Three days later they're staring at a carcase they glued up before fitting the drawers, and the opening is 1/32" tight, and the only fix involves a saw.
Sequence mistakes are the most expensive mistakes in the shop. Not because they're dramatic. Because they're irreversible.
Surface first. Always.
Before anything gets cut, the stock gets surfaced. S4S. Four flat faces, four square edges. This isn't about tidiness. It's about reference.
A router table fence referenced against a twisted edge gives you a dado in the wrong place. A mortise laid out from a concave face angles away from your line. When you surface first, every operation after it has something true to register against. When you skip it, you're stacking assumptions on top of each other and hoping they cancel out. They don't.
Joinery before shaping
The mortise goes in the leg before you taper it. The dovetail shoulders get cut on square stock before the profile work starts. This is geometry, not bureaucracy.
Joinery setups need the squared ends and flat faces that shaping removes. That's reason enough. But here's the real one: a mistake at joinery on square stock costs you one part. A mistake at joinery on a tapered leg with a molded foot costs you four hours of shaping you'll never get back.
Surface. Square the ends. Cut the joinery. Shape. In that order. Not approximately.
Sub-assemblies and the access question
Most plans show you the finished piece. They don't tell you what to glue up first. Getting it wrong locks you out of things you still need to fit.
The principle is simple: you can only adjust what you can still reach.
Leg-and-apron table. Two opposite leg pairs get glued first, one apron each. Then the two end assemblies get pulled together with the long aprons. Why? Because a four-corner glue-up all at once means open time working against you, clamps fighting each other, and no way to check square before things set. Two-at-a-time is manageable. Four-at-once is punishment.
Drawer boxes before drawer fronts. The box gets fitted in the opening first, while you can still reach through the back of the carcase to adjust the slides. Once the face frame goes on, that access is gone. So the face frame goes on after the drawers fit. Not before.
Same logic for anything inside the carcase. Shelf pins, blocking, wire management. If your hand needs to get in there, it needs to get in there before the thing closes.
Points of no return
Every build has moments where access shuts. The back panel goes in. The top gets attached. The face frame gets glued. Past those points, adjustment becomes reconstruction.
Before you start cutting, walk through the plan and ask one question: what do you lose access to as this piece comes together? Write it down. That list is your build sequence.
Before the back panel: pre-finish the interior surfaces. Drill the shelf pin holes. Run any wire channels.
Before the face frame: fit every drawer and door in the carcase openings. You can still shave from behind.
Before the top: install any hardware that needs access from above.
Before any sub-assembly glue-up: finish the interior faces. Getting a brush into a fully assembled inside corner is not woodworking. It's a penalty for not thinking ahead.
Pre-finishing: the trap
The sequence inside the sequence. Finish the surfaces that will be unreachable once assembled. Interior carcase walls. The inside of a frame-and-panel door's grooves. Drawer sides before the bottom slides in.
But joinery surfaces don't get finished. A tenon cheek needs bare wood for the glue to hold. A floating panel can't have oil in the groove where it needs to move freely.
The protocol: mask the joinery surfaces. Sand and finish the field. Pull the tape. Glue. It takes fifteen minutes. It saves the kind of regret that shows up six months later as a finish failure you can't reach.
Reading it out
The plan is geometrically complete. It is not sequentially complete. That's not a flaw. A drawing shows space. Sequence is time. Different media, different jobs.
You read the sequence out yourself. Look for sub-assembly boundaries. Trace which assembly closes off which surfaces. Follow the fitting dependencies: the door fits the opening, the opening is part of the carcase, the carcase has to exist first. Every part answers one question: what has to exist before you can make this?
The plan is the argument. The sequence is how you prove you understood it.