Capacity and Delivery Data Inputs

Capacity and Delivery Data Inputs are the measured, historical pieces of information about a team's actual capacity and past delivery performance that feed into release planning, providing an empirical foundation for forecasting rather than relying on assumption or optimism. They translate a team's demonstrated track record into concrete figures usable for projecting how much work can realistically be delivered across a release horizon.

Core Concept

Grounding Forecasts in Evidence

Release plans that rely purely on stakeholder hope or unvalidated estimation tend to be unreliable. Capacity and Delivery Data Inputs replace that reliance with measured data drawn from the team's own recent history, producing forecasts that reflect what the team has actually demonstrated it can achieve rather than what it might achieve under idealized conditions.

Two Related but Distinct Categories

Capacity data describes what the team is theoretically available to work, drawing on the calculations discussed throughout team capacity planning, while delivery data describes what the team has actually completed historically, which may differ from raw capacity due to estimation error, unplanned work, or process inefficiency.

Key Capacity Inputs

Effective Team Capacity

The calculated availability of the team after accounting for absence, overhead, and reserved buffers, providing the theoretical ceiling for what could be planned in each sprint within the release horizon.

Skill Distribution Data

Information about which skills are available in what quantity, feeding into release scope decomposition and sequencing decisions to avoid planning work the team lacks the specific expertise to complete on schedule.

Shared Resource Allocation

Data on how much of any shared specialists' time is actually available to the team, rather than nominally allocated, which directly affects how much release scope depending on those specialists can realistically be planned.

Key Delivery Inputs

Historical Velocity

The amount of work, typically in story points or completed items, the team has finished per sprint over a rolling window of recent iterations, widely used as the primary proxy for expected future delivery rate.

Velocity Variance

The degree of fluctuation observed in velocity across past sprints, used to establish appropriate confidence ranges around release forecasts rather than presenting a single deterministic number as though it were certain.

Unplanned Work Ratio

The historical proportion of sprint capacity consumed by unplanned or interrupting work, informing how much of the release horizon's total capacity should be reserved rather than allocated to planned scope.

Estimation Accuracy Trends

Data comparing estimated effort to actual effort for completed items, revealing systematic biases, such as consistent underestimation of certain categories of work, that should be corrected for in forecasting future release scope.

Using These Inputs in Release Planning

Establishing a Realistic Release-Level Capacity Baseline

Combining historical velocity with known upcoming changes, such as a confirmed new hire or a planned departure, produces a more accurate projection of capacity across the release horizon than relying on either historical data or current headcount alone.

Where $a_s$ is an adjustment factor reflecting known changes to capacity expected in sprint $s$.

Setting Confidence Ranges Rather Than Single Estimates

Applying observed velocity variance to release-level forecasts produces a range of plausible completion dates or achievable scope, communicating appropriate uncertainty to stakeholders rather than false precision.

Validating Proposed Scope Against Historical Throughput

Comparing a proposed release scope's total estimated effort against what historical delivery data suggests is realistically achievable serves as a direct input to the capacity plan validation process.

Risks of Poor or Missing Data Inputs

Forecasts Anchored to Aspiration Rather Than Evidence

Without reliable capacity and delivery data, release forecasts default to optimistic assumptions that rarely hold under actual execution, producing a pattern of chronic release delay.

Ignoring Variance Produces False Confidence

Presenting a release forecast as a single fixed date without incorporating known historical variance misleads stakeholders into expecting a precision the underlying data does not support.

Stale or Insufficient Historical Data

Using data from a period no longer representative of the current team, such as before a significant composition change, produces forecasts based on outdated assumptions rather than present reality.

Best Practices

Maintain a Clean, Consistent Historical Record

Tracking velocity and delivery data consistently across sprints, using a stable definition of what counts as completed work, ensures the resulting historical dataset remains a reliable basis for forecasting.

Use a Rolling Window, Not All-Time Averages

Basing forecasts on a rolling window of recent sprints, rather than the team's entire historical record, keeps the data responsive to current conditions rather than diluted by outdated performance.

Communicate Uncertainty Alongside Point Estimates

Presenting release forecasts with explicit confidence ranges derived from historical variance, rather than a single number, sets more accurate expectations and preserves stakeholder trust when actual results vary within that range.