Composite, drawn from patterns across engagements, not one specific client.
A project shipped six weeks late. The first instinct, from leadership, was that the developer had underdelivered. Looking at the actual timeline changed that read completely.
What the timeline actually showed
Development work itself finished only nine days behind the original estimate, reasonable slippage for a project of that size, not the source of six weeks of delay. The other five weeks were spent waiting: three separate rounds of stakeholder feedback that each took over a week to arrive, and one dependency on a third-party API integration that required approval from a partner company that took over two weeks to respond to a single email.
The build itself was never the bottleneck. The waiting between decisions was.
Why the wrong story stuck initially
Development is the visible, trackable part of a project, there's a task board, there's a person's name attached to it. Waiting on external feedback or approval doesn't show up on the same board, it's invisible time, and invisible time gets silently attributed to whoever's work is visible instead of to the actual cause.
What changed once the real timeline was visible
Feedback rounds got a hard 48-hour SLA going forward, with an explicit escalation if missed. The third-party dependency got flagged and initiated in week one of future projects instead of being discovered as a blocker mid-project. Neither fix touched development speed at all, because development speed was never the problem.
The lesson, generalized
"Late" gets attributed to whoever's work is visible on a timeline, almost by default. The actual cause is disproportionately often the invisible waiting between decisions, not the execution itself. Before concluding a delay is a delivery problem, check how much of the timeline was actually spent waiting on someone else, and who that someone else was.