How to set baselines, targets and forecasts for project benefits

Define the measure before agreeing the numbers

A baseline, target and forecast are useful only when people understand what is being measured. Describe the intended improvement, the population or activity covered, the unit, the observation method and whether higher or lower is better.

Check the time basis as well. An average processing time for completed cases in a month is different from the age of all cases still open at month end. Similar labels can conceal different questions.

APM describes benefits management as including definition, planning, tracking and realisation. This guide focuses on a practical part of that work: making the starting point and expectations clear enough to support a later review.

Keep four values distinct

A target is a commitment or agreed objective. A forecast is an expectation that should reflect current evidence. They can be equal, but they perform different jobs. Making a forecast more realistic does not automatically revise the target.

An actual is an observation, not automatically proof that the initiative caused the change. Retain the context needed to interpret it, including material changes in population, method or external conditions.

Establish a usable baseline and dated target

Use a starting observation that is relevant to the intended improvement. Record its source and period, and explain significant limitations such as seasonal demand or an unusual service interruption. A convenient single value may be a poor representation of the starting position.

If a defensible baseline is unavailable, make that gap visible and arrange the observation needed. Do not relabel an aspiration or an unsupported estimate as an observed starting point. The interim review may need to make a narrower claim until the gap is resolved.

Agree the target with the people responsible for delivering and using the improvement. Include the date, relevant scope, assumptions and any quality condition that must be preserved. Faster processing is not useful if it creates unacceptable errors or excludes difficult cases.

Build the forecast and review the comparisons

Use a forecast path that reflects the expected timing of delivery, adoption and operational change. An immediate improvement or a straight-line progression should be an explicit assumption, rather than the default shape chosen because it is easy to draw.

Use five checks when preparing the review.

1. Confirm that baseline, target, forecast and actual refer to compatible definitions and scope.

2. Compare the completed period’s actual with the relevant starting position and that period’s forecast.

3. Compare the forecast at the target date with the agreed target.

4. Investigate missing observations and material differences, taking account of whether higher or lower is better.

5. Record the decision, action or further evidence needed, including any agreed change to expectations.

A result can match the current forecast while the forecast still falls short of the target. Conversely, a result can improve on the baseline while being behind the expected pace. Keep both questions visible rather than relying on one favourable indicator.

Test a target gap with a simple example

In a fictional service programme, the observed baseline is an average processing time of 40 minutes. The agreed target is 25 minutes by the end of June, with the same case coverage and quality expectations.

At the March review, the actual is 34 minutes and matches the March forecast. The current June forecast is 30 minutes. March is performing as expected, but the expected June position still misses the agreed target.

The team needs to examine the assumptions and options. It might change the enabling work, improve adoption or seek agreement to revise the commitment. Simply changing the June forecast to 25 would conceal the current expectation; silently changing the target to 30 would conceal the original agreement.

The values illustrate the comparison only. They are not a customer result or a specification of a software status formula.

Preserve the meaning when using Wovex

Wovex supports benefits measurement, period-based tracking, collection and review. Demonstrate the intended measure with a baseline, dated target, period forecasts and completed-period actuals, then inspect the comparison used by the proposed review output.

Confirm where each agreed value is recorded and which comparison the output reads. Do not assume a field labelled Original Forecast is automatically the agreed target, or that recording a target in one place populates every reporting comparison.

Keep forecast changes formally agreed and explainable through the decision record and relevant period context. Confirm the current plan, configuration and supported workflow. Useful records and notes preserve the explanation; they should not be described as automatic snapshots or immutable history.