Quality is not a gate at the end of the line
A team that only tests at the end is not doing quality. It is doing damage assessment.
The first thing I ask when I join a quality area is simple: at what point does someone first say "this might not work?"
If the answer is "when it reaches QA", the area is not doing quality. It is doing damage assessment — careful, well documented, and far too late to be cheap.
The cost curve nobody draws
A problem caught while the requirement is being written costs a conversation. The same problem caught in homologation costs a rework, a re-test and a renegotiated date. In production it costs all of that plus the customer.
Nothing in that curve is controversial. What is hard is acting on it, because acting on it means quality people spending time in rooms where nothing is testable yet.
Prevention looks like boredom
A preventive quality area produces fewer heroic saves. There is no dramatic catch the night before release, because the ambiguity that would have caused it was resolved three weeks earlier by someone asking what happens when the field comes in empty.
That is a hard thing to put on a dashboard. It is still the whole job.
Where the gate still belongs
None of this removes homologation. It changes what it is for: confirming a decision that was already made on purpose, instead of discovering one that was never made at all.
