Too often, EMC is treated as something to verify after the layout, enclosure, cabling, and firmware are already fixed. By that stage, the available corrective actions are usually limited, expensive, and disruptive.
A better approach is to review EMC risk progressively during development. That means looking early at switching-current loops, reference-plane continuity, cable and connector boundaries, enclosure seams and apertures, grounding strategy, and even firmware operating states before layout and tooling are frozen.
Pre-compliance measurements can then be used not simply to ask “does it pass?” but to understand why an emission exists: is the dominant issue the noise source itself, the coupling path, or the physical structure that is acting as the antenna?
I’d be interested to hear how others structure these EMC design gates in practice. At what stages do you review EMC risk, and which measurements or design reviews have been most effective in preventing late-stage compliance failures?