Managing Delivery Performance: Lead Time, On-Time Delivery, Quality

You manage delivery performance through three coupled metrics on the same work package: how long it takes, whether the date holds and whether the result is right. Any single metric invites distortion, because whoever improves only speed buys it with defects, and whoever protects only dates builds buffers that lengthen the flow. Only a joint view shows whether an organization delivers reliably. This article lays out definitions and formulas and proposes a management rhythm for the executive team.
Why a single metric steers in the wrong direction
A unit whose head is measured on lead time shortens it fastest by skipping review steps. A unit measured on on-time delivery names only dates it is certain to meet and thereby lengthens the recipient's wait. A unit measured on zero defects adds extra approvals and loses speed. Each of these reactions is rational from the unit's point of view. Taken together, they produce an organization that looks good on every metric yet still reaches the recipient late or with defects.
Steering therefore needs a metrics triangle in which each figure limits the distortion of the other two. The precondition is a shared definition of “delivered”. From a delivery perspective, a work package is delivered when the recipient has accepted it and can use it. A completion mark in the project plan is not enough.
What is delivery management, and what does a delivery manager do?
Delivery management is the steering of delivery across teams and initiatives, with the aim that committed results reliably reach the recipient. A delivery manager owns this flow: they measure the delivery metrics and make sure that commitments rest on sound data.
The role is clearly distinct from neighbouring roles. A project manager owns a single initiative until acceptance. A product owner decides what a product delivers next. The delivery manager owns how reliably delivery happens and works on dependencies and bottlenecks between teams to that end. In organizations with several teams working on shared outcomes, it is the role that makes delivery capability visible as a whole. The article characteristics and metrics of a high-performing organization describes how the role fits into structure and steering.
How should lead time be measured?
Lead time measures the span from the commitment to a work package until its acceptance and is reported as a distribution with the median and the 85th percentile. An average hides the slow cases by which recipients judge the organization.
In knowledge work, the greater part of this span is typically spent waiting: for an approval, for feedback or for free capacity. Flow efficiency makes this share visible. If it is low, faster processing achieves little, and shorter waiting times achieve a great deal.
Little's law from queueing theory provides a formal frame. In a stable system, average lead time over a longer period equals average work in progress divided by average throughput. At constant throughput, lead time falls only if work in progress falls. Capping started work and starting new initiatives only when others are finished shortens lead time without extra capacity.
What should on-time delivery be?
Set the target per type of delivery and derive the commitment from the matching percentile of past lead times: if you want 85 percent on-time delivery, commit to the 85th percentile. For commitments to customers a target between 85 and 95 percent is advisable; for internal deliveries, 85 percent is a sound starting value.
On-time delivery is the share of work packages accepted by the committed date. It is complemented by the size of the deviation, because a delay of two days has a different effect from one of two months. The commitment already decides on-time delivery: setting dates out of wishful thinking or negotiating pressure creates delay in advance.
The target is a business decision for the executive team. A higher target means longer commitments or more capacity, and moving from 85 to 95 percent usually costs far more commitment time than moving from 70 to 85, because the upper percentiles of a lead-time distribution lie far apart. It therefore belongs in the target agreement between the executive team and the divisions.
On-time delivery should be distinguished from delivery reliability in the broader sense. On-time delivery asks only whether the date was met. Delivery reliability also asks whether scope and quality matched the commitment. Keeping the two apart is useful for steering, because it shows time and quality separately and so exposes the trade-off.
Quality: the counterweight to speed and dates
Quality is defined differently in every delivery setting, but the underlying question stays the same: how often does work have to be redone after acceptance? Suitable measures are the rate of acceptances without correction and the returns per delivery. Rework hours relative to total output complete the picture. The point of measurement matters: defects found before acceptance cost time, and if they appear afterwards the damage also hits the recipient's trust.
For software delivery, the DORA metrics (DevOps Research and Assessment, a research program running for more than ten years, now part of Google Cloud) provide a proven framework. Deployment frequency, change lead time and failed deployment recovery time stand for throughput; change failure rate and rework rate stand for stability. The research shows that high-performing teams achieve high throughput and high stability at the same time. The principle transfers to any type of delivery: for every speed metric, a stability metric belongs in the same report. The article Measuring team performance describes how DORA metrics can be used at team level.
A management rhythm for the executive team
The metrics take effect in the rhythm of decisions. At team level, work in progress and lead time are reviewed weekly, because bottlenecks can be fixed there immediately. Divisional leadership reads on-time delivery and the quality rate per type of delivery monthly and decides on capacity and priority. The executive team reviews targets and commitment practice quarterly.
Each metric has a threshold agreed in advance; when it is crossed, a named role decides. An example: if the 85th percentile of lead time rises by more than a fifth for two consecutive months, divisional leadership decides on limiting new initiatives. These decision paths are part of organization design.
Niehues Consulting as a partner for the build-up
Niehues Consulting builds this steering together with the people accountable. As delivery manager, project manager or product owner, Marin Niehues takes over the steering of running initiatives and sets up metrics and escalation paths along the way. Organization design defines roles and decision paths so that the organization delivers reliably in day-to-day business and in transformation. For AI initiatives, management of the AI portfolio with metrics is added.
The basis is the Niehues Delivery Model with its phases Understand, Build and Roll out. Three review questions matter most for steering delivery performance:
- What is the baseline, and how do we measure success? The starting point is the lead-time distribution and today's on-time delivery.
- Who owns operations from day one? Every metric has a named role that decides at the threshold.
- Which metric shows after 90 days that the solution is landing? Usually it is a falling 85th percentile of lead time with a stable quality rate.
The entry point is a free 30-minute initial consultation.
Frequently asked questions
What is the difference between lead time and cycle time?
Lead time runs from the commitment to a work package, cycle time from the start of work. The difference is the waiting time before the start. Because the terms are used inconsistently in practice, fix the definition in your reporting.
What is a good on-time delivery rate?
A good on-time delivery rate matches the target the executive team sets per type of delivery, for customer commitments between 85 and 95 percent. The commitment is derived from the matching percentile of past lead times, so for 85 percent on-time delivery from the 85th percentile.
What is the difference between on-time delivery and delivery reliability?
On-time delivery measures whether a date was met; delivery reliability also measures whether scope and quality matched the commitment. Tracked separately, the two show whether an organization meets dates at the expense of quality.
How do you calculate flow efficiency?
Flow efficiency is active work time divided by lead time, multiplied by 100. Three days of work in a lead time of 20 days gives 15 percent; on the remaining days the case is waiting.
What does Little's law state?
Little's law states that in a stable system average lead time equals average work in progress divided by average throughput. It follows that less work started at the same time shortens lead time without extra capacity.
What does a delivery manager do?
A delivery manager owns the reliable arrival of committed results with the recipient. The delivery manager steers the delivery metrics, removes obstacles between teams and ensures commitments are based on data.
Which metrics belong in a delivery report for the executive team?
A delivery report contains lead time, on-time delivery and a quality rate, complemented by work in progress as a leading indicator. Each metric needs a definition, a target and a threshold at which a named role decides.
Let’s talk about your initiative.
In 30 minutes we clarify where your organization stands and which step will pay off most now.
Request a consultation30 minutes · free · no obligation



