4 min read

Stop overwriting the plan: show baseline and forecast on the same Gantt chart

Stop overwriting the plan: show baseline and forecast on the same Gantt chart

Your delivery date has moved.

The status meeting is tomorrow. You open the Gantt chart—and the old date has vanished because the schedule was updated to match the latest forecast.

The chart now looks tidy, but it cannot answer the question everyone will ask: what changed?

A useful schedule readout needs to preserve two different truths. The baseline is the approved plan used for comparison. The forecast is the current estimate of what will happen given actual progress and what remains. If every update replaces the baseline, variance disappears from the picture even though the delivery risk remains.

The fix is simple in concept: show baseline, actual progress, and forecast on the same Gantt view. Then connect the visible variance to a decision.

First, separate the three layers

The U.S. Government Accountability Office’s schedule guidance treats a baseline schedule, actual progress, schedule variance, and an integrated schedule as management controls (GAO-16-89G). The guide is written for federal acquisition work, so it does not establish a universal chart format or review cadence. But its control logic is useful well beyond that setting: preserve the approved reference, record what has happened, and forecast the remaining work using valid schedule logic.

On one Gantt chart, display three layers:

  1. **Baseline bars:** the approved start and finish dates. Keep these visually stable so readers can see the reference plan.
  2. **Actuals through a named status date:** the work that has really started or finished as of a specific date. The status date is the boundary between observed progress and forecast work.
  3. **Forecast bars:** the current start and finish estimates for incomplete work, calculated from current progress, remaining duration, dependencies, calendars, and constraints.

These layers answer different questions. The baseline asks, “What did we approve?” Actuals ask, “What has happened?” The forecast asks, “What do we now expect?” Combining them does not remove uncertainty. It makes the uncertainty discussable.

Build the readout around a decision

A comparison chart is only useful if it helps someone decide. Add four fields beside the schedule:

- Variance reason: the evidence-backed explanation for the gap, not a color alone.

- Affected path or constraint: the dependency chain, resource limit, calendar, or milestone that makes the variance consequential.

- Decision owner: the person or role authorized to choose among the available responses.

- Next decision date: the point at which waiting would remove an option or increase exposure.

This turns “the milestone is red” into a decision frame. A reader can see the approved intent, the current outlook, why they differ, and who must act.

Do not calculate variance from a moving reference. If an approved baseline finish is Friday and the current forecast is Tuesday, compare Tuesday with Friday. Updating the forecast is routine schedule maintenance. Changing the baseline is a governance action whose approval and documentation should follow the organization’s own controls. There is no universal threshold that says when a baseline must change.

A worked example

Hypothetical example: A team plans a production cutover for Friday. Its schedule uses a Monday-to-Friday work calendar and no weekend work. A rehearsal planned for Wednesday uncovers a configuration issue. The fix and a repeat rehearsal consume two working days, moving the current cutover forecast to Tuesday.

The team keeps Friday as the baseline finish and draws Tuesday as the forecast finish. It records the reason as “configuration correction and repeat rehearsal,” identifies the rehearsal-to-cutover path as affected, names the release owner as the decision owner, and sets Thursday as the next decision date.

The schedule review can now compare two responses:

- Keep the planned cutover scope and move the forecast to Tuesday.

- Test whether a smaller cutover can still satisfy the required release outcome on Friday.

The chart does not choose between those options. It gives decision-makers a shared set of dates and assumptions. If the smaller cutover would compromise a required control, the Friday option should be rejected. If Tuesday conflicts with another operational constraint, that conflict belongs in the decision record.

The example is deliberately simple. Real schedules may include multiple calendars, resource leveling, lags, hard constraints, and converging paths. Those conditions should be modeled rather than hidden behind a manually moved bar.

Five steps for your next status readout

### 1. Name the status date

Write the “as of” date on the chart. Without it, readers cannot tell where actual progress ends and forecast assumptions begin.

### 2. Freeze the approved reference

Display baseline start and finish bars as the comparison layer. If the baseline has been formally changed, keep the change record and identify which approved version the chart uses.

### 3. Update actuals and remaining work

Record real starts and finishes through the status date. Re-estimate incomplete work from current evidence. Check that dependencies, calendars, and constraints still describe how the work is expected to flow.

### 4. Explain material variance

For each variance that changes a decision, write the reason and the affected path or constraint. “Late” describes a condition; it does not explain it.

### 5. Attach an owner and decision date

State who can decide and when the choice must be made. A variance without an owner becomes a recurring status observation.

Common mistakes

Silently replacing the baseline. This makes the current forecast look like the original commitment and destroys the comparison.

Treating the baseline as a promise that can never be governed. A baseline is a control reference, not a reason to ignore an approved change in scope or strategy. Preserve the history and use the organization’s change process.

Showing dates without schedule logic. A visually precise forecast can still be unreliable if dependencies, calendars, resources, or constraints are wrong. GAO’s guidance emphasizes the quality of the integrated schedule, not the appearance of a chart (GAO-16-89G).

Explaining every small movement. Focus the readout on variance that affects a milestone, constraint, option, or decision. The right level of detail depends on the cost of being wrong and the audience’s authority.

Using color as the explanation. Red, amber, and green can direct attention, but they cannot replace a variance reason, an assumption, or a decision.

What this view can—and cannot—do

A baseline-versus-forecast Gantt chart can make change visible and improve the quality of a schedule conversation. It cannot prove that the forecast is accurate, establish causation, or choose a response. Those depend on valid source data, schedule logic, judgment, and governance.

GAO’s guide is federal acquisition guidance. This article adapts its control principles into a practical visualization; it does not claim that this exact template or any particular review cadence improves delivery in every organization.

For your next readout, keep the approved baseline on screen. Add actuals through a named status date, forecast the remaining work, and label the variance reason, affected decision, owner, and decision date. The goal is not to make the schedule look stable. It is to make change visible early enough to manage it.

Source

- U.S. Government Accountability Office, *Schedule Assessment Guide: Best Practices for Project Schedules*, GAO-16-89G: https://www.gao.gov/assets/d1689G.pdf