The epiphany of integration points is that they control product development and are the leverage points to improve the system. When timing of integration points slips, the project is in trouble.
—Dantar P. Oosterwal
Principle #4 – Build incrementally with fast, integrated learning cycles
In traditional, phase-gated development, investment costs begin immediately and accumulate until a Solution is delivered. Often, little to no actual value is provided before all of the committed Features are available, or the program runs out of time or money. During development, getting any meaningful feedback is challenging because the process isn’t designed for it. Moreover, the development process itself isn’t set up or implemented to allow incremental capabilities to be evaluated by the customer. As a result, the risk remains in the program until the deadline and even into deployment and after initial use.
No wonder the process is error-prone and problematic, often resulting in a loss of trust with the customer. Attempting to adjust for this, both parties try even harder to define the requirements and select the best design up-front. They also typically implement even more rigorous phase gates. Each of these remedies, unfortunately, actually compounds the underlying problem. This is a systems-level problem in the development process: it must be addressed systemically.
Faster Learning through Faster Cycles
The better approach begins with establishing a systematic way to accelerate learning. As originally defined by Shewhart and advanced by Deming  , the plan-do-check-adjust (PDCA) cycle is the mechanism for gaining faster feedback and controlling the variability of solution development . In many aspects, PDCA is simply an expression of the scientific method for creating a hypothesis, experimenting, and evaluating the results to navigate uncertainty and create new learning. It is an iterative cycle, as illustrated in Figure 1. Regardless of the results, every experiment creates new insights that help form the following hypothesis for learning to be gained or uncertainty to be resolved.
Fast feedback accelerates the acquisition of knowledge. Shorter PDCA cycles improve learning efficiency by decreasing the time between action and effect. They reduce the cost of risk-taking by truncating unsuccessful paths quickly. In other words, the shorter the cycles, the faster the learning. Smaller batch sizes and modern, automated development environments are two key moves that enable shorter cycles.
Integration Points Control Product Development
Establishing shorter PDCA cycles enables more frequent integration points. In complex systems development, local integration points are used to ensure that each system element or capability meets its responsibilities to contribute to the solution Intent. These local points must then be integrated at the next higher system level, as shown in Figure 2.
These integration points accelerate the learning process. However, development cannot proceed any faster than the slowest learning loop. Synchronizing the learning loops enables faster and more frequent integrations.
The larger the system, the more such integration levels exist. Solution designers recognize that the top-level, least-frequent integration point provides the only true measure of system progress and work to create these points as frequently as possible. All stakeholders understand that when the timing of integration points slips, the project is in trouble. But even then, this knowledge helps spark the necessary adjustments to scope, technical approach, cost, or delivery timing needed to redirect the effort to meet revised expectations.
Integration Points Reduce Risk
Any lack of integration points creates a false sense of security. This results in the belief that linear progress is being made when the problems are hidden in reality, only to be discovered later. This creates what Oosterwall  terms as ‘false positive feasibility,’ i.e. the belief that the solution will be delivered as planned, whereas, in reality, that is not the case, as Figure 3 illustrates.
Lean principles and practices approach the problem differently. Rather than pick a single requirements-and-design choice early on—assuming that it’s feasible and will provide fitness for purpose—a range of requirements and design options (Principle #3 – Assume variability and preserve options) are considered while building the solution incrementally fast, integrated cycles. Each results in an increment of a working system that can be integrated and evaluated. Subsequent cycles build on the previous increments, and the solution evolves until it’s released. The knowledge gained from integration points is not solely to establish technical viability. Many integration points reduce risk by providing minimum viable solutions or prototypes for testing the market, validating usability, and gaining objective customer feedback. Where necessary, these fast feedback points allow teams to pivot to an alternate course of action that should better serve the needs of the intended customers.
Integration Points Occur by Intent
The development process and the solution architecture are designed, in part, to focus on cadence-based integration points. Each point creates a ‘pull event’ that pulls the various solution elements into an integrated whole, even though it addresses only a portion of the system intent. Integration points pull the stakeholders together, creating a routine synchronization that helps assure that the evolving solution addresses the real and current business needs instead of the assumptions established at the beginning. Each integration point delivers its value by converting uncertainty into knowledge:
- Knowledge of the technical feasibility of the current design choice
- Knowledge of the potential sustainability of the solution based on objective measures (Principle #5 – Base milestones on objective evaluation of working systems)
 Oosterwal, Dantar P. The Lean Machine: How Harley-Davidson Drove Top-Line Growth and Profitability with Revolutionary Lean Product Development. Amacom, 2010.
 Ward, Allan C., and Durward Sobek. Lean Product and Process Development. Lean Enterprise Institute Inc., 2014.
 Deming, W. Edwards. Out of the Crisis. MIT Press, 2000.
Last update: 28 November 2022