The Schedule Was Healthy Until We Counted Humans

Filed by: D.B. Trench

Field line: A plan that ignores capacity is just optimism with gridlines.

The schedule looked fine in the deck.

Four workstreams. Two weeks of testing. One week of defect resolution. A tidy launch readiness checkpoint. The kind of plan that makes executives relax because the bars are straight and nobody has asked the spreadsheet whether people sleep.

The testing lead did not relax.

She had stopped making jokes.

That is a metric I trust more than most dashboards.

When a testing lead is still joking, the project may be ugly but air remains in the system. When she goes quiet and starts answering in nouns, check the capacity.

So we counted humans.

Not roles. Not named resources in a table. Actual available hours.

The plan needed 258 hours of testing and defect retest over ten working days. Confirmed tester availability was 147 hours. The business reviewers had another 36 hours between them, assuming no sick leave, no urgent operations work, no meeting drift, and no one discovering that a test script had been written by a person who believed in miracles.

The project was short before it began.

The first version of the capacity note was too polite:

Testing availability may create pressure on the current schedule.

May create pressure is what people write when they are trying to warn the building without waking the residents.

The second version did the math:

The current test plan requires 258 hours of execution and retest. Confirmed availability is 147 tester hours and 36 business-review hours. The schedule has an evidence gap of 75 hours before defect turnaround, meetings, and operational interruptions. Without scope reduction, added capacity, or date movement, testing will proceed with incomplete coverage.

That sentence changed the conversation.

It did not fix the project.

That distinction matters.

Leadership did not add capacity. The date did not move. Two lower-priority scenarios were deferred, and the support window was expanded because reality did not care about our preference.

The testing lead still carried too much. The team still worked late. The project still absorbed a portion of the risk it had just named.

But the record changed.

No one could say testing was simply “challenging.” No one could imply that incomplete evidence came from poor coordination. The math was there before the pain arrived.

At launch readiness, the note was still uncomfortable:

Testing completed against reduced coverage path. Deferred scenarios and expanded support window are accepted launch conditions due to confirmed capacity gap.

That was not a victory speech.

It was a receipt.

If the schedule looks healthy but people are getting quiet, count hours in the open. Required effort. Confirmed availability. Skill constraints. Review time. Retest time. Decision time.

Then do not confuse the math with authority.

The math had arrived early. The decision arrived late.

That is how capacity failures usually happen. Not because nobody counted. Because counting was treated as commentary until the people became the buffer.

---

For the next time the bars look calm and the people go quiet:

Capacity reality check: Use this when the dates are approved but the available people and hours have not been counted.

Required: [hours/skills/reviews/retest/decision time].

Available: [confirmed people/hours], with known operational interruptions stated plainly.

Decision required: Convert the gap into added capacity, reduced scope, date movement, or accepted risk.

Record line: “Coverage/date confidence is conditional on closing the [hour/resource] gap by [date].”

Related kit: The Delivery Survival Bundle is built for projects where the plan looks fine until the people are counted.