The wedge is a decision
Industrial software is often introduced as a system of record. The more useful starting point is usually a recurring decision that is currently slow, fragile, or dependent on one experienced operator.
A focused decision creates a measurable before and after. It also gives the product a reason to collect context that customers already care about.
Context should compound
The defensible layer is rarely a dashboard. It is the practical model of a site: its constraints, exceptions, operating rhythm, and consequences.
Products become harder to replace when that model improves the next decision without increasing administrative work.
Questions to carry forward
What decision becomes materially better? What context is earned through normal use? Which adjacent workflow becomes obvious once the first decision works?