Why Most Construction Projects Finish Late

Why Most Construction Projects Finish Late

Here’s the deal. Most construction projects are late before anybody sets a single piece of steel. Not late because the crews were slow, and not late because the weather turned. Late because the schedule they were handed was never possible in the first place, and everyone signed it anyway.

That is a hard thing to say out loud in a preconstruction meeting. It is harder to live through in month fourteen, when the superintendent is getting asked why drywall is behind and the honest answer is that the plan never had room for a single bad day.

You Are Behind in Week One

You know the feeling. The durations on the bar chart do not match what your hands know about the work. You walk the zone and there are four trades in a space that comfortably holds one. The electrician is roughing in overhead while the framer is still setting track below him, and the mechanical contractor is staging pipe in the only clear lane. Nobody is doing anything wrong. There is just too much work stuffed into too little space and too little time.

Or you get the opposite version of the same problem. Your best trade partner is asked to be in Zone 2 on the third floor and Zone 5 on the first floor in the same week, with one crew. So they split the crew, both areas move at half speed, both fall behind, and the schedule still shows both on track, because a bar chart does not know that people cannot be in two places at once. Then the pressure starts. Look ahead meetings get tense, the owner asks for a recovery plan, everyone works weekends, and the project still slides.

The Failure Is in the Method

Here is the pattern underneath all of it, and it is not a people problem. It is a method problem. Most projects carry the critical path method as a contractual requirement, and CPM was designed with a specific intent, which was to find the fastest possible duration for a project. That sounds reasonable until you look at what it actually produces. What CPM produces is not the fastest path. It produces the most impossible path.

Look at what has to be true for a CPM bar chart to close. The overall duration gets compressed to its most unreasonable length. To make that math work, the schedule stacks trades, which means too many crews in one area at one time, or it burdens trades, which means asking a single trade to be in multiple areas at once. There is no float on the critical path by definition, and there are no buffers anywhere, because buffers look like waste to somebody optimizing a bar chart. So you have a plan with no room to absorb a delay, on a jobsite where delays are guaranteed. That is not a schedule. It is a bet.

I want to be clear about something, because field teams carry shame that does not belong to them. The system failed them. They did not fail the system. The superintendent who cannot hit the duration in week one is not underperforming, and the foreman who cannot cover two zones with one crew is not uncommitted. Every one of those people is being measured against a plan that was mathematically impossible on the day it printed, then blamed for the gap, and then they go home tired and convinced they are the problem. They are not the problem. The method is, and the method is changeable.

Why CPM Projects Crash at the Two Thirds Mark

Watch what happens on a CPM project when reality shows up. A delay hits. Then another delay hits. Then another. Each one eats into a duration that had no slack to give, while the contractual end date has not moved, because the contract does not care about your logic.

So the owner starts breathing down the neck of the project delivery team, and the team has exactly one move left inside CPM theory. They start dissolving logic at the end of the schedule. Activities that used to be tied together get untied. Work that clearly has to happen after other work is now shown happening at the same time, because that is the only way the bar chart closes on the contract date. The schedule looks fixed, and nothing about the actual work got easier.

That is why CPM projects crash land at the two thirds mark. It is not bad luck and it is not coincidence. Somebody on the delivery team was told to dissolve that logic, and the pretend float created at the end of the schedule finally runs out right where the work is densest and the trade count is highest. Every project we analyze with CPM shows the same shape. The dissolved logic. The crash at two thirds. The finish that happens through heroics, overtime, and burned relationships instead of through flow.

What It Actually Costs

Signals your schedule has already gone impossible:

• More than one trade assigned to the same zone in the same period without a documented handoff
• The same trade named in multiple zones during the same week with one crew
• Zero float shown anywhere on the driving path
• No named buffers between phases, only activity durations

Those are not scheduling details. They are the leading indicators of a project that will finish late, over budget, and with a worse safety record than it should have had. Trade stacking is a safety problem before it is a schedule problem, because it means people working over and under each other and rushing. Quality suffers next, since rework is what happens when trades cannot get clean access and adequate time. Then the money goes, through acceleration claims, overtime premiums, and the slow bleed of low productivity that nobody itemizes.

Then it reaches the part that matters most. Missed dinners. Saturday shifts that turn into Sunday shifts. A foreman who stops mentoring his apprentices because he is too busy expediting material. A superintendent who starts to believe that chaos is simply what construction is. That cost shows up in no report, and it is the largest cost we pay.

Build the System, Then Build the People

So what do we do instead? Two things, in this order. We use the right system, and we build the people who run it. Projects fail because of a lack of planning and the use of a method that cannot hold a plan together. The alternative is Takt planning, and Takt is not a faster bar chart. It is a different unit of thinking. Instead of asking how short each activity can be, Takt asks how work should flow through space at a steady rhythm. You define zones. You define a Takt time, which is the fixed duration a crew spends in a zone before handing it off. Then you build a train of trades that moves through those zones in sequence at that beat. Stacking goes away because each zone belongs to one trade for one Takt, and burdening goes away because a crew is assigned to one zone at a time.

Then you add what CPM refuses to include, which is buffers. Real buffers, placed on purpose, named in the plan, with rules for who can consume them and when. Buffers are not padding, and they are not slack a scheduler hides inside durations. They are the shock absorbers that let a project take a hit and keep its rhythm. A plan with buffers absorbs a late submittal or a bad weather week without dissolving its own logic. A plan without buffers has no response except pretending. The duration a Takt plan produces is reasonable, which is the entire point, because a reasonable duration the team can hold finishes earlier than an impossible one that collapses at two thirds. Slowing the individual activity speeds up the project. Tools like LeanTakt exist to keep that flow visible.

Now the second thing, and this is the one people skip. We do not train our people. We build things before we build people. A Takt plan handed to an untrained team becomes a wall chart nobody follows by month three. The superintendent needs to know how to run a Takt board, hold a handoff, protect a buffer, and say no to a request that would burden a trade. The foremen need to know what a clean handoff looks like and what to do when a zone is not ready. That is a training investment, not a software purchase. If your project needs superintendent coaching, project support, or leadership development, Elevate Construction can help your field teams stabilize, schedule, and flow. Most teams do not need to be told they are behind. They need somebody to build the system with them and stay long enough that it holds after the coach leaves.

This Is About the People on the Jobsite

None of this is really about schedules. It is about what a jobsite does to the people standing on it. A stacked, burdened, buffer free project treats craft workers as interchangeable capacity to be pushed around a bar chart. A Takt project treats them as skilled people who deserve a clean, ready area, a defined amount of time, and a predictable week. That is what dignity looks like in construction. It looks like a foreman who knows on Monday what Thursday holds, and an apprentice who learns properly because the journeyman was not sprinting.

When teams make this change, crews report predictable weeks before the schedule shows improvement, handoff conversations get shorter and less defensive, and trade partners start bringing problems forward earlier. Flow is the mechanism, stability is the result, and respect is the reason.

So here is the challenge. Pull your current schedule this week and look for the three tells. Find the zones holding more than one trade. Find the trades named in more than one zone at once. Find the buffers, and if there are none, you have your answer about how this project ends. Do not wait for the two thirds mark to confirm it.

Edwards Deming said it plainly: “A bad system will beat a good person every time.” Your people are good. Give them a system worth their effort.

On we go.

FAQ

Is CPM ever useful on a construction project?
CPM works as a contractual and analytical artifact. Where it fails is as a production system, because it has no mechanism for flow and no place for buffers. Many teams keep CPM for the contract and run production with a Takt plan underneath it.

What is the difference between trade stacking and trade burdening?
Trade stacking is too many trades in one area at the same time, which creates congestion, safety exposure, and rework. Trade burdening is asking one trade to be in multiple areas at once, which splits crews and slows both areas. CPM schedules produce both, because compressing a duration requires overlapping work that cannot actually overlap in real space with real people.

Why do CPM projects fail specifically around the two thirds point?
Because that is when the borrowed time runs out. Delays accumulate early, and the only response available inside CPM is to dissolve logic at the end of the schedule so the bar chart still closes on the contract date. That creates float that does not exist, so the project arrives at its densest, most trade heavy stretch with no reserve, and the plan collapses.

Where should a team start if they are mid project on a failing CPM schedule?
Start with zones and a single trade train through the remaining work, then place buffers ahead of the milestones that matter most. Do not try to convert the whole schedule at once. Take the next three months, make the flow visible, train the superintendent and foremen on how to run the handoffs, and let the results earn permission for the rest.