Essay

The Real Obstacle

Hard programs do not usually fail because nobody is working. They fail because the work is not converging on the constraint that determines success.

Frank GarciaEngineering judgment

In a difficult engineering program, activity is easy to see. People are designing boards, writing software, running tests, holding reviews, updating schedules, and solving hundreds of legitimate problems. Yet the program can remain strangely motionless.

The usual response is to add urgency: more meetings, more tracking, more people, more parallel work. Sometimes that helps. Often it amplifies the motion without improving the direction.

The central leadership question is not “Are we working hard?” It is “What single obstacle, if left unresolved, makes the rest of this work unable to converge?”

Symptoms are persuasive

The wrong obstacle is often attractive because it is visible and locally solvable. A firmware bug, a late board, a test failure, or an understaffed function becomes the center of attention. But those can be symptoms of an unclear interface, an impossible requirement, an unowned integration boundary, or a development sequence that asks for evidence before the system can produce it.

Integration reveals truth

Components can each appear healthy while the system is failing. Integration forces hidden assumptions into the open: timing, tolerances, ownership, operating modes, fault behavior, manufacturing variance, and what “done” actually means.

This is why experienced technical leadership often looks less like supplying answers and more like restructuring the questions. What must be true? What evidence would prove it? Which risk can invalidate the rest of the plan? Who owns the boundary? What is the shortest path to learning?

Work the constraint

Once the real obstacle is named, the program becomes easier to organize. Work can be sequenced around retiring the decisive risk. Teams can distinguish enabling tasks from merely available tasks. Reviews can examine evidence rather than presentation quality. The schedule becomes an expression of technical reality instead of optimism.

The result is not necessarily comfort. The real obstacle may be difficult, expensive, or politically inconvenient. But it creates something more valuable than comfort: convergence.