Book Notes: Edwards Deming's "Out of the Crisis" (Ch 1 through 3)
"Unfortunately, quality control departments have taken the job of quality away from the people that can contribute the most to quality [...]."
In Book Notes, I quote passages from a book with my thoughts and analysis beneath the passage on how these ideas can be applied to improve the design and construction industry.
When I first stepped into quality role, I searched for books and other resources to learn about not just quality in construction, but the origins of quality itself. Edwards Deming was one of the fathers of quality or “quality gurus”. His book, Out of the Crisis, helped me look at not just quality but business operations differently. This text is considered one of the foundations of quality for all businesses.
The notes below were the most influential for me. Note that page numbers are from the hardcover edition, fifteenth printing (1991).
This post covers chapters 1, 2 and 3. Book Notes for future chapters are forthcoming in February.
Here’s a preview from Chapter 3.
Pg 133: "Unfortunately, quality control departments have taken the job of quality away from the people that can contribute the most to quality [...]."
This is the case with centralized quality programs. The dedicated quality manager does all the inspections, facilitates all quality-related meetings, and takes on most of the quality responsibility. Those who are installing the work have the greatest impact on quality. This requires that they are part of all preconstruction and preinstallation meetings, and understand the client’s intent. The quality manager can be a facilitator of this process, but not execute the actual work. (This doesn’t apply when a quality manager is required contractually in public or mission critical projects.)
When responsibility for any work effort - schedule, safety, quality, cost - is centralized, projects become overstaffed. Organizations over-hire. For example, a company may hire too many superintendents and a quality manager. By spreading the quality workload across the superintendents for the work they manage, the quality manager would not be required. And, having a streamlined program that the superintendents could understand would reduce the number of needed superintendents. Larger project teams creates more friction for collaboration, which increases quality and safety incidents.
Complicated quality systems require a quality manager to decipher the work instructions for superintendents and project managers. These create significant overhead expenses for an organization. This is relevant for dedicated scheduling staff and safety staff.
Actionable Takeaways
Shift Quality Responsibility Back to Those Doing the Work
Quick Win: Require trade partners and field teams to attend all preconstruction and preinstallation meetings to ensure they understand the client’s intent.
Long-Term Strategy: Train superintendents and project managers to integrate quality checks into daily operations rather than relying solely on a dedicated quality manager.
Chapter 1
Pg 2: "When we improve quality we also improve productivity."
This aligns with the idea that we manage for quality (more on this here). Quality can’t be managed by itself as a separate process or system (more on this here). Quality and productivity are results.
This also ties to safety. Safety and quality need to be looked at together, as they both impact the other. A study cited by ENR noted that most safety incidents occur during rework - fixing poor quality work (here’s a link to the original ENR post).
Pg 5: "Quality begins with the intent, which is fixed by management."
This was the most important point in the book. I included Client Intent as a source of a Definable Features of Work due this statement (more analysis can be found here on creating a list of DFOWs, using Client Intent to understand what a client really wants out of the project).
We need a tangible, objective means for talking about what the client wants beyond asking "What's important to you?” In construction and design, we focus on being under budget, on time, and safety, thinking that if we achieve those three objectives, we’ve also achieved quality.
Pg 6: "The problem is where to find good management."
Has management defined what “quality” or “success” is? Do they know what good looks like?
What about the people who work for them?
Most managers don’t realize that improvement is their responsibility - not that of production.
Pg 8: "This is an example of a gain in productivity accomplished by a change in the system, namely, improvement in definitions, effected by management, to help people work smarter, not harder."
A major problem with quality in the industry is that its definition is not tactical for those installing the work.
There is also no agreement or definition of quality within an organization (more on how to define quality in your company can be found here).
We need to simplify our quality systems and terminology so that everyone on a project can understand what’s expected of them (more on simplifying quality here). Again, this is management’s responsibility.
For example, does everyone on the team know the difference between quality assurance and quality control, and what tasks are associated with each?)
Pg 11: "We now had an identifiable process."
Can your teams identify the quality process?
If they can identify the steps, do they know what is required in each step?
Pg 11: "Defects are not free. Somebody makes them, and gets paid for making them."
What does it cost for someone to install work incorrectly because they were not aware of what drawing to install from?
Pg 12: "In some work I did for a railway, study showed that mechanics in a huge repair shop spent three-quarters of their time waiting in line to get parts."
Across your company, how often are your team members waiting? Identify these "waiting" points as places to start with process improvement.
Pg 13: “Lag in American productivity has been attributed in editorials and in letters in the newspapers to failure to install new machinery, gadgets, and the latest types of automation such as robots.”
Interesting that back in 1982, they were having the same conversations around "technology" as we are today.
In theory, our modern work products should be perfect, given the technology we have - but they aren't. Why is that?
Technology has saved us so much administrative time, yet many teams still complain they don’t have time for a particular task.
There is no perfect software or technology (I wrote more on this here).
Automation and simplification are separate tasks. We must simplify first, automate last (I wrote more on this here).
Actionable Takeaways
Define Quality and Productivity as a Management Responsibility
Quick Win: Leadership must clearly define what "quality" and "success" look like in tangible, tactical terms for every project.
Long-Term Strategy: Establish a company-wide understanding that quality is not an isolated process—it is directly tied to productivity and safety, requiring a holistic management approach.
Align Quality with Client Intent and Clear Definitions
Quick Win: Move beyond the standard “on time, under budget, safe” approach and define what the client actually values in their project.
Long-Term Strategy: Implement a structured process for defining Definable Features of Work (DFOWs) based on client intent, ensuring teams align their work with these expectations.
Remove Process Inefficiencies That Lead to Rework and Waste
Quick Win: Identify common “waiting points” in your projects—whether it’s delays due to missing materials, unclear drawings, or ineffective communication—and address them immediately.
Long-Term Strategy: Conduct regular workflow assessments to systematically reduce wasted time, rework, and inefficiencies, ultimately improving productivity and safety.
Simplify Quality Systems Before Automating Them
Quick Win: Audit your current quality systems to remove unnecessary complexity before introducing new software or automation tools.
Long-Term Strategy: Develop a streamlined, easy-to-understand quality management framework that ensures all team members—especially those installing the work—understand their role in delivering quality.
Shift the Conversation on Technology—It’s a Tool, Not a Solution
Quick Win: Stop assuming that new technology will automatically fix productivity and quality issues. Instead, focus on process improvements first.
Long-Term Strategy: Train teams to use technology as a support system rather than a crutch. Prioritize simplification before automation to avoid reinforcing inefficient processes with expensive software.
Chapter 2
This entire chapter has a lot to offer. Highlights are below.
Pg 18: "The transformation can only be accomplished by man, not by hardware."
"If we only had xyz technology / software, we could collaborate better together." See my previous comments on technology from Chapter 1.
Technology will help us, but it won’t solve our problems.
Pg 21: "'[I]f you don't have time to do your job, there is not much I can do for you."
Yes, time is a serious limitation we need to account for, but we can only do so much.
It’s not that we don’t have time, rather that we decide to work on the wrong tasks, or we take on too much. There is no time management system that will enable us to get everything done. Some tasks won’t ever get done. We decide what must get done while being comfortable not getting other work done. (I wrote more on this here.)
Pg 22: "[S]omeone in management, walking around, has little idea about what questions to ask, and usually does not pause long enough at any spot to get the right answer."
Many leaders are too far removed from the construction work to have an impact on improving the system, yet they are often the only ones with the authority to do so.
Pg 23: "Cease dependence on inspection to achieve quality." And, pg 29: "You can not inspect quality into a product."'
See my Book Notes for Atul Gawande’s The Checklist Manifesto.
Many in our industry believe quality is the inspection effort. “Yes, we have a quality process. We inspect the work.” Quality is larger than that. Quality involves all the planning efforts before the works starts as well.
The inspection is also too late to ensure the work is installed correctly. If the inspection catches incorrect work, we’ve already paid someone to install it incorrectly.
Pg 30: "Divided responsibility means that nobody is responsible."
The opposite also occurs. When we assign a dedicated quality manager to a project, they are often the only one responsible for quality.
This is why a decentralized quality system may be what the industry needs. (I wrote previously on decentralizing quality here.)
Pg 31: "A better way, in my opinion, would be to improve the forms for clarity and ease; also to introduce training to help the typists to understand what constitutes a mistake and its consequences."
We need to be more clear about what good looks like.
We also need to simplify all our forms ("forms" includes meetings, checklists, all our documentation) without compromising the intent of the tool.
Quality programs have become defensive, which creates an administrative burden for project teams. (I wrote previously about defensive processes here.)
Pg 49: "[...] quality must be built in at the design stage."
Once construction starts, it’s too late to improve the quality of a procured material or product.
A robust and collaborative design process also drives quality as a result in design. The client’s intent must be highlighted in the design documents, and the Definable Features of Work clearly outlined for construction. Using the DFOW terminology in design is untraditional, however, we should all be using similar language. Those installing the work may not understand what design intent means, or may not interpret it the same way. And, is design intent what the client wants?
Pg 49: "We repeat here that the quality desired starts with the intent, which is fixed by management."
I’ll repeat again as well: "[...] all of which are management's responsibility."
Pg 59: "Secure means without fear, not afraid to express ideas, not afraid to ask questions."
Many in your organization may be too intimidated to tell you the truth about your quality program. (I wrote previously about the two ways a quality leader can act here - be a collaborator.)
Pg 63: "Each staff area was suboptimizing its own work, but not working as a team for the company."
Design “suboptimizes” its own work, as does construction. The two phases need to optimize together. To accomplish that, we need to view the project in the same way. (See my previous article on DFOWs and Legos.)
Pg 63: "Management often complicate the work of the people in design by making last-minute changes in style and engineering, after the plans are submitted and production is ready, leaving the design and production engineers only a few weeks to do a year's work."
Designers still live this every day. If we use Definable Features of Work to help everyone understand where to put their focus, this will help reduce the potential for last minute change.
How often does a change happen because we learn something new that has to be incorporated?
How many changes occur because your client didn’t understand the project’s design? Many clients do not construct new facilities very often. I’ve met clients who are not able to look at drawings and visualize what is being built.
Pg 67: "The management needs to learn that the main responsibility is theirs from now on to improve the system."
Only when leaders take on responsibility for system improvement can the responsibility of quality be put on production.
Pg 69: "[...] numerical goals set for other people, without a road map to reach the goal, have effects opposite to the effects sought."
We ask that project team members conduct x number of safety or quality inspections, or that they take y number of photos.
Do teams know why they conduct inspections? Or that inspections are in fact too late to have an impact on quality?
Pg 76: "A goal beyond the capability of the system will not be reached."
Do your current systems - if team members are fully aware of them, an obstacle itself - allow for better deliverables?
Does your current quality program have the capacity to deliver tangible results?
We can't set goals when we've lost track of the process.
Has your company outgrown the capacity of its processes?
Pg 76: "Focus on outcome is not an effective way to improve a process or an activity."
Without a clear definition of what quality is, those managing and installing the work do not know how to achieve it.
A common desired outcome is to “build per plan,” which assumes that all team members are able to interpret the client’s intent from the drawings. This is not always the case.
Pg 96: "What architect would be so completely oblivious to the customer? One was. What purchaser of the building would have the same failing? One did."
But why was the architect oblivious? We must always assume positive intent. Why didn't the process include a means by which the architect spent time to truly understand what the client wanted?
Actionable Takeaways
Technology is a Tool, Not a Solution
Quick Win: Stop relying on technology as a silver bullet for collaboration and efficiency. Instead, focus on improving communication and processes first.
Long-Term Strategy: Invest in training teams to optimize existing systems before layering in software. If a process is broken, technology will only make bad work happen faster.
Time Management is About Priorities, Not More Hours
Quick Win: Reevaluate daily tasks and cut unnecessary work. Not everything needs to get done—only the highest-value tasks.
Long-Term Strategy: Shift project planning toward prioritization rather than just time tracking. Leaders must guide teams in making trade-offs between what must be done and what can be left out.
Quality Can’t Be Inspected In—It Must Be Built In
Quick Win: Shift from relying on inspections to focusing on pre-planning and early collaboration in design and construction.
Long-Term Strategy: Develop a decentralized quality system where responsibility is shared across the project team rather than assigned solely to a quality manager.
Align Design and Construction to Prevent Last-Minute Changes
Quick Win: Improve clarity in design documents to minimize misinterpretations by the construction team.
Long-Term Strategy: Use Definable Features of Work (DFOWs) to create a common language between design and construction, ensuring all stakeholders understand client intent.
Management Must Take Responsibility for System Improvement
Quick Win: Leaders must actively seek honest feedback about quality programs and be open to process improvements.
Long-Term Strategy: Shift from setting arbitrary numerical targets (e.g., inspections completed, photos taken) to focusing on whether teams truly understand and execute quality work.
Define Quality Before Setting Goals
Quick Win: Clearly define what “quality” means at every level of the project, from design to execution.
Long-Term Strategy: Ensure processes are capable of achieving quality goals before setting them. If a system is outdated or overloaded, no goal will be met effectively.
Chapter 3
Pg 97: "[...] cure of deadly diseases will require total reconstruction of Western management."
The point here is we need to completely rethink how we manage our work, processes, and people before any improvement can happen. I’ve suggested breaking up the project into manageable chunks - Definable Features of Work - for both design and construction as a means to improve productivity and quality.
Pg 98: "[...] continually toward improvement of processes and of product and service that will bring the customer back again."
Improvement is full time and will never end. Years from now, we will still be improving the construction quality process. We don't know everything to the degree to have only one improvement effort. Markets and people change, requiring constant adoption and improvement.
Pg 102: "[...] caused totally by the system that they work in."
It's a fault of the system, not the people.
If your projects are not achieving the intended outcomes, what is the process by which they are designed and built? Are they simple systems and processes? Easy to use? Clear for anyone to understand?
Pg 105: "Counts relieve management of the necessity to contrive a measure with meaning."
We count the number of inspections, meetings, and other items. We define good as having met a number or having held a meeting.
Instead, we should measure the understanding of the project amongst the project team. Has everyone read the specifications? (Many don’t.) Are the specifications applicable to the project? (Many include items that are not.)
Pg 109: "Don't work with those people [in another group]." Your time belongs to our project."
How can we share the knowledge? (I’ve written more on lessons learned and transferring knowledge here.)
Many organizations institute mentoring and training programs for which people don't have time to attend because their time always "belongs" to another budget or project.
Pg 126: "It is not so simple: it will be necessary to study and go to work."
We need to think critically and learn - to struggle through new knowledge and ideas - before we can improve.
Training programs are one-time efforts to improve performance. Short-term improvements. The long-term solution is creating processes with steps that trigger the exchange of tacit knowledge.
Pg 133: "Unfortunately, quality control departments have taken the job of quality away from the people that can contribute the most to quality [...]."
This is the case with centralized quality programs. The dedicated quality manager does all the inspections, facilitates all quality-related meetings, and takes on most of the quality responsibility. Those who are installing the work have the greatest impact on quality. This requires that they are part of all preconstruction and preinstallation meetings, and understand the client’s intent. The quality manager can be a facilitator of this process, but not execute the actual work. (This doesn’t apply when a quality manager is required contractually in public or mission critical projects.)
When responsibility for any work effort - schedule, safety, quality, cost - is centralized, projects become overstaffed. Organizations over-hire. For example, a company may hire too many superintendents and a quality manager. By spreading the quality workload across the superintendents for the work they manage, the quality manager would not be required. And, having a streamlined program that the superintendents could understand would reduce the number of needed superintendents. Larger project teams creates more friction for collaboration, which increases quality and safety incidents.
Complicated quality systems require a quality manager to decipher the work instructions for superintendents and project managers. These create significant overhead expenses for an organization. This is relevant for dedicated scheduling staff and safety staff.
Pg 144: "It becomes a work standard: meet it, but don't beat it."
Defensive, cumbersome processes and work standards don’t allow for anyone to exceed the standard. Time that could be used for exceeding standards is instead used up trying to understand the process and expectations.
Actionable Takeaways
Rethink Management Before Improvement Can Happen
Quick Win: Break projects into Definable Features of Work (DFOWs) to create manageable chunks for both design and construction.
Long-Term Strategy: Shift from reactive management (fixing problems as they arise) to proactive systems that simplify work, clarify expectations, and drive continuous improvement.
Continuous Improvement is Non-Negotiable
Quick Win: Set up regular, structured project debriefs to capture lessons learned in real-time.
Long-Term Strategy: Build a culture where quality and efficiency are never considered “finished” but are always evolving in response to market and operational shifts.
Problems Are Systemic, Not Individual
Quick Win: Evaluate project systems and workflows—if results aren’t meeting expectations, focus on redesigning the process rather than blaming individuals.
Long-Term Strategy: Simplify procedures to be intuitive and accessible. The best systems are easy to follow, reducing errors and misinterpretations.
Stop Measuring the Wrong Things
Quick Win: Instead of tracking the number of inspections, meetings, or reports, measure understanding—has the team read and grasped the specifications? Are they relevant to the project?
Long-Term Strategy: Move from superficial compliance metrics to deeper measures of effectiveness, such as how well knowledge is transferred within the team.
Decentralize Quality, Scheduling, and Safety Responsibilities
Quick Win: Instead of relying solely on a quality manager, involve superintendents and project teams in preconstruction and preinstallation meetings to share the responsibility.
Long-Term Strategy: Reduce the need for excess staff by creating streamlined, decentralized systems where responsibility is embedded into roles rather than concentrated in a single department.
Overly Rigid Standards Limit Excellence
Quick Win: Identify where cumbersome quality and safety processes are slowing progress and look for ways to simplify without compromising intent.
Long-Term Strategy: Shift from defensive compliance-based processes to frameworks that allow teams to exceed standards rather than just meet them.
Thanks for reading! You can find other Book Notes here.