Stakes fail in one of two ways. They are absent, in which case the piece is pleasant and weightless. Or they are enormous and abstract — the fate of the world, the future of the industry — which is functionally the same as absent, because nobody can picture it. Specific and small beats vast and general in almost every case.
The strongest stakes are things the audience has already been shown the value of. This is why setup matters so much: five seconds establishing that the workshop is the only thing he has left does more for the later scene than any amount of dialogue about how much it means to him.
In commercial work the honest stakes are usually modest and that is fine. Someone will have to explain a missed deadline. Someone will lose the thing they were good at. Those land because they are recognisable, which is a property "transforming the industry" does not have.
What are stakes in storytelling?
What the protagonist stands to lose if they fail. Effective stakes are specific enough that the audience can picture the loss, which is why small and concrete usually outperforms vast and abstract.
How do you raise the stakes without making them absurd?
Narrow them rather than enlarging them. Show what the thing at risk is worth to this particular person, early and briefly, so the later threat lands without needing to be inflated.