Baselines and the Control Loop

What control means

Control is not supervision. It is a loop: measure where the project actually is, compare it against where it should be, understand the difference, decide what to do, act, and measure again.

Each step is routinely skipped. Projects measure without comparing, compare without analysing, analyse without deciding, and decide without acting — producing reports that describe difficulty without changing anything.

Baselines

Control requires something to control against. The three baselines are scope, schedule and cost, together forming the performance measurement baseline.

A baseline is approved, frozen and changed only through change control. Its purpose is to preserve the original commitment so that variance is visible. Where a plan is continuously updated without a baseline, the project is always on track by definition, because the plan moves to match whatever happened.

When a baseline is changed through approved change control, keep the history. A project reporting on time against its fourth re-baselined schedule is not on time, and the sequence of baseline changes is itself the most honest summary of what has happened.

Tolerance

Governance should define the range within which the project manager proceeds without escalation — typically expressed as a percentage or absolute variance on time and cost, plus defined scope limits.

Tolerance makes control workable. Without it, either everything is escalated, which paralyses governance and the project, or nothing is, which means problems surface only when they are large.

Frequency

Control cadence should match the pace of the work and the consequence of being wrong. A fast-moving project needs weekly measurement; a long infrastructure project may need monthly. The test: could something go significantly wrong and not be noticed until the next cycle? If so, the cycle is too long.

What to measure

Control means integrated measurement, not separate silos:

• Scope — what has been delivered and accepted against what was planned.
• Schedule — progress against the baseline and the forecast completion.
• Cost — spent, committed and forecast.
• Quality — defects, rework, test results.
• Risk and issues — the changing exposure.
• Resources — availability against requirement.
• Benefits — whether the expected value is still credible.

Measuring these separately conceals the relationships that matter. A project on budget is not reassuring if it is on budget because a third of the work has not been started.