Responsive design solved the layout problem well enough that many teams consider mobile handled. The interface reflows, nothing overlaps, and the checklist is satisfied. Yet conversion on mobile stays materially lower, and it is not because of the width.
The context is different, not just the screen
Desktop use is typically deliberate, seated, and uninterrupted. Mobile use happens in queues, on transport, between other tasks, frequently one-handed and often on a slow connection.
A form that is merely tedious on desktop becomes an abandonment on mobile, because the user has a genuine alternative: stopping.
Design the smallest version first, honestly
Starting from the constrained case forces a decision about what genuinely matters, and that decision usually improves the desktop version too. Starting from desktop and removing things produces a mobile experience defined by what would not fit.
Every field costs more than it looks
Typing on a phone is slow and error-prone. Each optional field, each free-text box, and each field that could have been inferred is a real cost paid by every user.
The correct keyboard type, sensible autofill, and defaults that are right most of the time do more for mobile conversion than most redesigns.
Test on a real mid-range device
A recent flagship on office wifi is not representative. A three-year-old mid-range phone on a mobile network is, and the experience on it is frequently several times slower than the version the team has been reviewing.
Respect the physical hand
Primary actions belong within thumb reach, targets need real size, and destructive controls should not sit next to the ones people press constantly. This is ergonomics, and it does not follow automatically from a responsive grid.
Mobile is where most traffic already is. Designing for it last means designing for most users last.