Iteration Forecast

Iteration Forecast is the projected estimate, produced at the conclusion of planning, of how much of the sprint backlog the team expects to complete by the end of the iteration, expressed with an honest acknowledgment of the uncertainty inherent in any such prediction. It differs from a simple commitment statement in that it explicitly frames delivery expectations as a probabilistic projection grounded in evidence, rather than an absolute guarantee, giving stakeholders a realistic basis for planning around the team's output.

Purpose of Forecasting

Setting Realistic Expectations

By presenting expected outcomes as a forecast rather than an ironclad promise, the team communicates that variability is a normal part of software delivery, helping stakeholders plan around a probable range of outcomes instead of anchoring rigidly to a single guaranteed figure.

Grounding Predictions in Evidence

A forecast is only as valuable as the data underlying it; iteration forecasting relies on historical velocity, current capacity, and known risks rather than optimism, making it a more trustworthy planning input than a purely aspirational commitment.

Supporting Downstream Planning

Teams and stakeholders responsible for coordinating dependent activities—such as marketing launches, customer communications, or downstream integration work—rely on the iteration forecast to plan their own timing with appropriate buffer.

Creating Accountability Without Rigidity

Because the forecast is explicitly probabilistic, it allows the team to be held accountable for reasonable, evidence-based prediction, while avoiding the dysfunction that arises when any deviation from a fixed commitment is treated as failure.

Inputs to the Forecast

Historical Velocity

The team's actual completed work from recent iterations forms the primary empirical basis for predicting what a similarly sized iteration is likely to deliver.

Current Sprint Backlog Composition

The specific mix of item sizes, complexity, and risk levels within the newly formed sprint backlog is compared against historical patterns to judge whether this iteration is likely to behave similarly to past ones.

Known Risks and Uncertainty

Findings from the risk and uncertainty review are factored into the forecast, since iterations containing significant unresolved risk warrant a more conservative prediction than iterations composed primarily of well-understood work.

Team Capacity Adjustments

Any known reductions in capacity, such as planned absences or anticipated support obligations, are incorporated to avoid a forecast based on an unrealistically full team.

Methods of Forecasting

Simple Velocity-Based Projection

The most straightforward approach compares the total estimated size of the sprint backlog against the team's average recent velocity to judge whether the committed scope is broadly consistent with demonstrated capability.

Range-Based Forecasting

Rather than predicting a single outcome, the team expresses a likely range, reflecting the natural variability observed across past iterations, giving stakeholders a sense of both the expected outcome and its plausible bounds.

Probabilistic Simulation

More mature teams may apply simulation techniques, such as Monte Carlo methods drawing on historical throughput variability, to generate a probability distribution over possible completion outcomes rather than a single point estimate.

Communicating the Forecast

Presenting Confidence Alongside the Number

Rather than stating a bare figure, effective communication pairs the forecast with an indication of confidence, such as "likely to complete between fourteen and eighteen items, based on the last six iterations," giving stakeholders a truer picture than a false sense of precision.

Updating the Forecast as the Iteration Progresses

As the iteration unfolds and actual progress becomes visible, the forecast can be refined, allowing stakeholders to receive updated, increasingly accurate expectations rather than only the original planning-time projection.

Explaining Deviations After the Fact

When actual outcomes diverge meaningfully from the forecast, reviewing why—whether due to unforeseen complexity, resolved uncertainty, or external disruption—helps refine future forecasting accuracy and maintains stakeholder trust in the process.

Common Failure Modes

Presenting the Forecast as a Guarantee

Communicating a probabilistic forecast using language that implies certainty misleads stakeholders and sets the team up for perceived failure when natural variability produces a different outcome.

Forecasting Without Historical Data

Attempting to forecast in the absence of sufficient velocity history results in guesses dressed up as forecasts, lacking the empirical grounding that gives forecasting its value.

Ignoring Known Risk in the Projection

Producing a forecast based purely on historical averages, without adjusting for unusually high risk or uncertainty in the current iteration's specific composition, can produce an overconfident and misleading prediction.

Treating a Single Missed Forecast as Systemic Failure

Reacting to one iteration falling outside its forecasted range as evidence of broad planning failure, rather than recognizing normal variability, can pressure teams toward artificially conservative forecasts that undersell their actual capability.