What changes when chocolate is made at scale

Grinding is the clearest split. A craft maker uses a melangeur — granite wheels over a granite bed, 24 to 72 hours, doing grinding and much of the conching at once. An industrial plant uses ball mills or roll refiners that hit target particle size in under two hours, with conching as a separate controlled operation.

That is not a shortcut. Separating the two operations gives independent control over texture and flavour, which the combined melangeur process does not offer.

Continuous tempering with inline crystal measurement. Roast profiling with feedback control. Particle size distribution measured rather than estimated. Consistency across thousands of tonnes and across years.

That consistency is a serious engineering achievement, and it is easy to disparage precisely because it succeeds so completely that nobody notices it.

Lot separation is the main casualty. An industrial line running continuously needs consistent input, which means blending across many farms and seasons, which means individual lot character disappears by design.

Roasting to reveal a particular bean is also impractical when the bean changes weekly. Dark roasting to a standard target is the rational response, and it is why industrial chocolate tastes of process rather than of place.

Craft chocolate at its best does things industrial chocolate structurally cannot — express a specific harvest from a specific farm. Craft chocolate at its worst is under-refined, inconsistently tempered and considerably worse than a competent mass-market bar.

Industrial chocolate is reliably decent and rarely remarkable. Which you want depends entirely on whether you are eating for interest or for a recipe.