When W. Edwards Deming wrote about quality, he wasn’t thinking about tilt-up panels, crane picks, or RFIs. Yet his 14 points map cleanly to the daily chaos of a construction site. I learned this the slow way, through change orders that should have been avoidable, punch lists that grew teeth, and schedules that turned into wish lists when the supply chain coughed. Deming’s thinking pulls you out of firefighting and into systems, which is where construction actually lives. Materials, methods, teams, contracts, and weather form a system. Manage the system, and the day-to-day follows.
The goal here is not to quote Deming like scripture. It’s to translate his principles into actions that make steel go up straighter, crews go home safer, and owners want to work with you again. I will keep the theory tied to the field, with a few missteps I’ve made and seen.
Constancy of purpose in a project-driven world
Construction tempts you to chase today’s crisis at the expense of tomorrow’s value. Deming’s first point, constancy of purpose, asks you to decide what you will be consistent about for the long haul. That isn’t a slogan. It is a small set of operational choices that outlast the bid phase and the latest PM assignment.
For a general contractor, constancy of purpose can be written into three targets that survive turnover: predictable safety performance, reliable schedule adherence, and right-first-time work. On one hospital build, we committed to a rework rate below 2 percent by value. That meant we slowed our submittal approvals early, huddled with trades to lock in connection details, and pretested building envelope mock-ups until the water stopped sneaking in. The owner barely noticed those early weeks. They noticed when we topped out on the date we wrote six months earlier and hit occupancy without a scramble.
Constancy also means investing in capabilities that do not pay back on a single job. Training foremen on look-ahead planning, building a standard commissioning checklist, and keeping a clean lessons-learned database are dull on paper but gold when crews rotate. If your priorities shift with every procurement officer change, you will chase your tail.
Adopt a new philosophy, not a new app
Many construction firms try to fix systemic problems with tools. Deming asks for a new philosophy: quality is built into the process, not inspected in at the end. In practical terms, that means managers stop rewarding heroic recoveries and start rewarding stable, boring flow.
I saw a framing subcontractor cut its average interior deficiency count by 60 percent, not by buying a scanner or a new punch app, but by agreeing on a common definition of ready with the MEP trades. They refused to start a stud wall until the slab was verified, sleeves were located, and hangers were laid out with chalk. Two weeks of grumbling, then the flow smoothed out. The app they already had simply became easier to use because there were fewer defects to capture.
A new philosophy also means seeing variability as the enemy. Sixteen different ways to seal a roof penetration guarantee surprises. Standardize the one that performs in your climate and on your budget, then teach it relentlessly.
Stop relying on end-of-line inspection
Inspections are essential, but they are not quality control. They are confirmation. Deming’s third point, cease dependence on mass inspection, is hard to swallow when authorities require hold points and sign-offs. The trick is to build control into the process so that inspections rarely discover anything new.
On a mid-rise residential project, we moved window installation quality upstream by building two full-scale mock-ups. The manufacturer sent their field tech. The crew practiced the sequence with the exact fasteners and tapes. We water-tested, found three failure points, adjusted the shim spacing, and trained the whole crew. When the city inspector arrived months later, he spent 15 minutes on windows and 45 minutes talking about sports. The point wasn’t to dodge scrutiny; it was to render it routine.
Quality at the source works for concrete as well. Instead of relying on cylinder breaks and a day-late panic, have the batch plant share moisture corrections live, use a real-time slump measurement device or at least a standardized field test protocol, and empower the finisher to reject a truck without drama. If rejecting a truck triggers paperwork purgatory, crews will take borderline loads and hope for the best. Set up the system so the right choice is the easy choice.
Don’t pick subcontractors by price alone
Deming’s fourth point, end the practice of awarding business on price tag alone, is almost offensive to procurement orthodoxy. Yet anyone who has lived through a “low-low” electrical bid knows the cost when the real price shows up in delays and claims. The smarter move is to prequalify, weigh past performance, staffing plans, and financial stability alongside the number on the proposal.
One public project I worked on had to take the lowest responsible bidder. That word responsible did a lot of work. We enforced prequalification on safety EMR, references from at least two similar projects, and proof of a superintendent with the hours and trade experience to handle the scope. The lowest bidder sometimes still won, but not the lowest in a vacuum. Over a three-year program, we saw claims drop by half. The owner noticed lower variability more than lower averages.
For private owners, best value selection feels obvious. Still, document the criteria and share them with prospective bidders. If a mechanical contractor knows you will assess their prefabrication capability and commissioning track record, they will show you their process, or they won’t bid. Either outcome is better than surprises after award.
Improve constantly and forever, but choose your few
Continuous improvement sounds like a poster in a break room. In construction, it becomes real when you clamp down the scope to a handful of improvement targets, measure them the same way across projects, and carry lessons forward without reinventing the wheel.
Pick three to five metrics: percent plan complete on weekly look-aheads, first-pass inspection acceptance, rework cost as a percent of contract value, RFI turnaround time, field change order frequency. Track them honestly. The first quarter is usually ugly, because reality is ugly. Then things get useful.
We ran kaizen events on drywall layout, crane picks, and document control. The best result came from a 90-minute huddle with carpenters who showed us how our submittal naming mess was costing them an hour a day. We rebuilt the folder structure, standardized the revision cloud callouts, and reduced drawing-related RFIs by roughly 30 percent over the next two months. No software purchase, just listening, and then sticking with the new method instead of sliding back.
Build training into the schedule
Deming’s sixth point, institute training on the job, is where construction often misses. We assume that craft workers arrive with full competence, and they often do in their trade. But every site is a new species. The sequence, the tolerances, the coordination surfaces, and the chosen products change job to job. If you treat training as optional, you pay through rework and schedule drift.
Training is more than a toolbox talk. For activities with high risk or tight tolerances, plan for start-up days. Schedule the manufacturer’s rep for the first run. Budget for the foreman to pull a small crew off the line to teach and observe. Include enough material for a practice panel, a mock-up wall, or a test batch. The time cost on day 4 is paid back by day 40.
This is especially true with technology. If you bought tablets and models, invest in field training that shows a carpenter exactly how to pull the door schedule and check a frame type without wading through a 500-page set. If the tablet is locked away in the office trailer, it is an ornament, not a tool.
Lead, don’t just count
Deming draws a hard line between leadership and supervision. Leaders remove obstacles so people can do quality work. In construction, obstacles are often invisible until a foreman points to them. A leader’s walk on site is not to catch people; it is to catch friction.
On a data center project, our best superintendent spent more time moving laydown areas than issuing directives. He saw that trades were hauling materials 200 feet farther than necessary because the initial plan assumed a different crane path. A half-day with cones and signage shaved minutes off every cart run, which equaled hours by the end of the week. He also met each trade’s lead at the end of day and asked the same question: what slowed you down today that I can fix by tomorrow? Not everything was fixable, but enough was to matter.
Leaders also interpret the contract to preserve relationships. Strict enforcement does not mean brittle behavior. If your steel sub is delayed by a drawing coordination hole you partially caused, acknowledge it and solve for impact, not blame. The money talk will be cleaner when trust is intact.
Drive out fear, so problems show up early
Nothing wrecks a project faster than a crew that hides mistakes. Deming’s eighth point, drive out fear, sounds soft until you measure the cost of truth delayed. On one office tower, a layout error on embeds was reported immediately, we assessed in hours, and a small base plate change solved it with no schedule hit. On a similar job years earlier, a crew waited two weeks, hoping to fix quietly; we ended up coring, epoxying, and negotiating for months.
You don’t drive out fear by being nice. You do it by responding predictably. When someone raises a problem early, treat the act of reporting as positive even if the content is negative. Document the issue, focus on containment and countermeasures, and reserve disciplinary conversations for willful negligence, not honest mistakes.
Anonymous safety reporting and clear near-miss logs help, but they are not a substitute for trust. If meetings turn into public scoldings, the logs will go quiet. Walk the talk by sharing your own misses. When a PM admits they misread a spec and shows the remedy, crews see that truth is safe.
Break down barriers between departments and trades
Construction divides itself by necessity: design vs. construction, office vs. field, trade vs. trade. Barriers become problems when handoffs are brittle. Deming’s ninth point is where building information modeling, pull planning, and co-location earn their keep. But tools don’t do the work unless people are in the same conversations.
We ran weekly design-constructability sessions with the architect, structural engineer, MEP coordinator, and at least one field superintendent. Half of the RFIs we thought we needed evaporated in those rooms. We also put the drywall foreman, electrician, and low-voltage lead together every Thursday for two months leading up to above-ceiling close-in. They marked out zones, agreed on hanger precedence, and pre-sequenced rooms. Above-ceiling inspection passed on first sweep in 80 percent of areas. In previous projects without those sessions, it was closer to 50 percent.
Cross-trade mock-ups work even when you don’t have BIM. Build a bathroom pod in the corner of the slab with actual trades. The act of touching the problem beats three weeks of email.
Eliminate slogans and empty targets
“Zero defects” posters don’t help a tired finisher who was sent the wrong mix at dusk with the pour already started. Deming’s tenth point tells us to stop setting exhortations that ignore the system. In construction, that means your metrics must be paired with means.
If you want fewer defects, give crews time to plan and inspect their own work before external inspections. If you want safety, adjust schedules so that overtime is the exception, not the base plan. If you want schedule reliability, protect look-ahead commitments from last-minute reprioritization. When you post a banner about quality, also show the change in your process that will make quality more likely.
A contractor I know replaced annual safety slogans with monthly problem-of-the-month focus. One month, they targeted ladder misuse and replaced a third of the aging ladders, trained on proper angles, and enforced tie-offs with grace for the first week. Incidents dropped not because people suddenly cared more, but because the system stopped setting them up to fail.
Remove quotas and arbitrary productivity targets
Task labor in construction often gets a quota by the hour or day: this many studs, that many fixtures. Deming’s eleventh point warns that numerical goals without method create cheating and stress. Crews meet the number by cutting corners or hiding issues.
This doesn’t mean we stop measuring productivity. It means we pair metrics with method and context. If the layout is clean, material is staged within reach, and interference is cleared, then a drywall crew can hit a known, fair production rate. If half the studs arrive dented and the lift is shared with another trade, the target becomes punitive.
Use takt planning or simple line-of-balance charts to visualize flow, not to threaten. When flow slips, investigate the constraint. Nine times out of ten, it is not laziness; it is a missing piece of the system.
Elevate pride in workmanship
Construction attracts people who like to point and say, I built that. Deming’s twelfth point, remove barriers to pride in workmanship, is about clearing the bureaucratic fog that blocks that feeling. The fastest way to kill pride is to bury crews in unclear drawings, impossible sequences, or constant rework from upstream mistakes.
Give trades clear acceptance criteria. Agree on tolerances, mock-ups, and inspection checklists before work begins, so success is visible and fair. Invite crews to punch their own areas with the architect before formal walks. The best crews take pride in finding and fixing their own misses; give them room and credit for it.
Recognition matters, but it should tie to craft, not just safety pizza parties. We started framing walkabouts where we praised clean corners, crisp reveals, and neat mechanical rooms. The tone shifted. People paid attention to craftsmanship visible to peers, not just schedule.
Institute education and self-improvement
Training teaches the task. Education raises the ceiling. Deming’s thirteenth point suggests investing in people beyond immediate job needs. In construction, that looks like sending a promising foreman to a scheduling course, teaching a PM basic estimating, or letting an engineer spend a week with the survey crew. Cross-training improves empathy and coordination. It also builds resilience when someone leaves or a project pivots.
On a long program, we rotated assistant superintendents through quality control for a quarter. They learned to read specs with a different eye and returned to the field more decisive. We also ran short lunch-and-learns with inspectors and code officials. When field staff heard directly why an inspector fixated on firestopping details, compliance climbed without the usual grumbling.
Education doesn’t need a classroom. Pair a younger engineer with a seasoned superintendent for a morning each week. Send carpenters to tour the millwork shop. Ask your concrete finisher to brief the design team about slab flatness, with examples. You accelerate trust when people understand each other’s domains.
Make transformation everyone’s job
Deming’s final point asks management to lead the transformation, but not to keep it to themselves. In construction, transformation dies when it becomes a corporate program that never reaches the slab. It lives when a plumber can explain how the new coordination process helps him avoid rework, and a project accountant can fix a pay app bottleneck because she owns that part of the system.
Start with one project as a laboratory. Define a few Deming-inspired changes: best value trade selection, field-first training on critical scopes, weekly cross-trade problem solving, and simple, stable metrics. Give the superintendent and PM license to adjust forms and meetings within reason. Visit the site and ask crews what actually changed for them. If the answer is new posters and more reports, you missed. If the answer is fewer surprises and more reliable days, you found traction.
Codify wins lightly. A one-page standard for above-ceiling coordination is better than a 60-page manual that no one reads. Roll changes to the next project but keep testing. Deming’s work is not a shrine. It is a habit.
Translating the 14 points to field practice
Readers often ask for concrete actions. The risk is turning Deming’s philosophy into a checklist divorced from context. Still, a few practical moves have broad utility when tuned to your project size and delivery method.
- Preconstruction discipline: bake constructability and quality into design with target value design sessions, early trade involvement, and decisive product selections. Freeze critical details earlier than your comfort zone and mock up assemblies with the actual installers. Scheduling for flow: use short-interval planning with weekly commitments, tie commitments to material readiness and access, and protect the plan by resolving constraints during daily huddles. Measure percent plan complete honestly and act on misses the same week. Quality at the source: create first-run studies for high-risk scopes, use mock-ups as training tools not trophies, empower trades to self-check against clear criteria, and treat third-party inspections as confirmation of a stable process. Best value procurement: prequalify trades on safety, financial health, and relevant experience; score proposals on means and methods; hold alignment meetings before award to surface assumptions; and right-size retainage to avoid cashflow strangulation. Culture that reveals problems: standardize rapid issue escalation paths, respond without theatrics, and hold weekly cross-trade coordination focused on barriers. Celebrate early problem reports with quick fixes more than flawless weeks that hide truth.
Contracts and delivery methods through a Deming lens
Contract structure can amplify or sabotage Deming’s principles. Lump sum often drives price-first selection and change-order jousting. Guaranteed maximum price can six sigma templates encourage early collaboration, but not if buyout is still a race to the bottom. Design-build aligns incentives, but only if the builder invites trade partners early and designers feel free to iterate without punishment.
You can make Deming work under any delivery method, but the tactics change. In hard-bid environments, invest more in prequalification and post-award alignment. Make preconstruction meetings about method, not only paperwork. In CM at risk, invite key trades into design earlier, share target values, and co-author details with those who will install them. Integrated project delivery goes furthest, but it demands adults in the room. Shared risk without shared trust becomes shared misery.
Regardless of form, write quality into contracts as process commitments, not only outcomes. Require first-run studies, mock-ups, and participation in pull planning. Tie payments to visible progress and verified quality, not just quantities. Give trades time and access for training, and back that with schedule logic.
Data and technology, the Deming way
Deming warned against tampering with systems based on noisy data. Construction technology now produces a blizzard of numbers: model checks, progress scans, safety observations, and sensor feeds. The temptation is to chase every blip. Resist. Choose a handful of meaningful indicators, understand normal variation, and act on signals, not noise.
For schedule reliability, track percent plan complete and reasons for misses. If late material delivery shows up as a recurring cause, that is a system issue you can attack. For quality, track first-pass inspection acceptance and categorize defects by root cause: design ambiguity, coordination miss, workmanship, or material. Only then decide whether to retrain, clarify, or change a supplier.
Use 3D models and reality capture to verify work in place against design, but close the loop. The value is not the pretty heat map; it is the coordination meeting where the survey shows anchors drifted and you adjust sequence before steel arrives. Keep tech close to the field. If your model lives with VDC in an office two states away, its value decays.
Safety as a system property
Deming’s philosophy fits safety perfectly. Incidents are rarely a single person’s fault; they arise from production pressure, unclear methods, poor housekeeping, or missing equipment. Measure leading indicators like participation in pre-task plans that are actually read and used, quality of hazard identification, and availability of the right tools.
I walked a site where housekeeping lagged. Instead of a crackdown, we added daily five-minute zone resets at 2 pm, scheduled dumpsters and brooms like any other resource, and rewarded trade crews for the cleanest area with first choice on Friday crane time. It became a game people wanted to win, and tripping hazards evaporated. We did not raise the bar; we moved the bar to where people could meet it.
Dealing with variability, especially supply chain
The last few years taught even stubborn builders that supply variability can swamp precise plans. Deming’s ideas push us to design systems that absorb variation without breaking. That means buffers placed thoughtfully, not everywhere.
Use time buffers at the end of critical chains, capacity buffers with flexible crews who can pivot to productive alternative tasks, and inventory buffers for long-lead items, stored safely and sequenced for access. Work with suppliers to get earlier signals of delays, not just delivery dates. Ask for leading indicators like factory capacity utilization or raw material constraints. Treat suppliers as partners with visibility, not vending machines.
On a school project, acoustic ceiling tiles became a joker. We adjusted the sequence to close walls and above-ceiling work first, pushed tile installation later, and secured partial shipments to finish classrooms in clusters rather than all at once. The school opened on time, with the gym finished last when the final batch arrived. The plan was ugly on paper, but it acknowledged reality and protected the milestone that mattered.
The human cadence of meetings
If your meetings feel like theater, your system is overloaded with ceremony. Deming would counsel parsimony: enough structure to see and solve problems, not so much that you create them. Daily huddles at the crew level, 15 minutes max, focused on what is ready, what is blocked, and who needs help. Weekly coordination that looks ahead two to three weeks, resolves constraints, and assigns owners with dates. Monthly executive reviews that check trends, not individual anecdotes.
Kill meetings that only repeat what software already shows. Move decisions to the lowest level consistent with risk. Share minutes that state decisions and assignments, not novels. The feel on site should be steady, not frantic.
Where Deming meets the realities of claims and disputes
Even with the best system, you will face claims. Deming’s philosophy does not ask you to be naïve. It asks you to document causes with clarity, prefer prevention over righteous anger, and preserve relationships for the next project. Keep contemporaneous records: daily logs, photos with dates, delivery receipts, and meeting notes. Use them to tell a calm story when an issue arises.
When you are wrong, own it early. When you are right, show it without gloating. Most disputes shrink when both sides can see the system breakdown rather than a villain. If a claim proceeds, the side with better process evidence usually prevails or settles well.
A short field checklist for Deming-minded builders
- Before mobilization: finalize three critical mock-ups and schedule manufacturer reps; prequalify trades on more than price; write first-run studies into the baseline. During ramp-up: hold cross-trade layout sessions; establish daily huddles and weekly constraint removal; agree on clear acceptance criteria for key scopes. In steady state: track percent plan complete, first-pass acceptance, and rework cost; publish reasons for plan misses; fix one systemic barrier each week. As turnover approaches: increase cross-discipline inspections; keep change discipline on owner-directed items; use a living punch list with trade-owned closeout tasks. After completion: hold a blunt lessons-learned session with field, office, and trades; update standards and details; recognize crews for craftsmanship with specifics.
The payoff you can bank on
When you run projects with Deming’s 14 points in mind, the benefits show up as lower variance. Fewer hair-on-fire weeks. Smoother inspections. Subcontractors who take your call on the next job because they were treated like partners. Owners who stop hovering because the data and the walk tell the same story. Your cost curve stabilizes, and your margins stop depending on miracles.
It is not magic. It is discipline, especially when the pressure mounts. The deming 14 principles are not a checklist to hang on the trailer wall. six sigma They are a way to design your construction system so that good outcomes are likely and bad surprises are rare. You will still have weather, design changes, and the occasional misstep. But you will have a team that sees problems early, fixes them without drama, and wants to build together again. That is project success, job after job.
