Why Construction Projects Always Go Over Budget 

🏗 Have you ever underestimated the cost of a project? Tell us the story below! 🪒 Look forward to your morning routine with a Henson razor (Code: PRACTICALENGINEERING for a free pack of 100 blades): bit.ly/3CWiWJP

When a contractor bids a realistic number, they're almost guaranteed to NOT win the project.

Civil Design Engineer here. I think it would do a world of good if civil design/project engineers actually had experience working at a construction site. There have been times during a design review where my manager asked me “How are they going to build this particular piece of the alignment?” I kinda shrug my shoulders and point out the reasons why I made that particular design choice and he usually agrees, but we always come back to how a contractor will build it. It really is my weakest part of my engineering knowledge. To all the future civil engineers, please do an internship with a contractor and learn from them. It will save you time in the future and imho, will be well worth it.

As someone that does construction estimating for a living I loved this video. Everyday is a constant battle to ride the fine line of expensive enough to make money but not too expensive to not get the job. When is really scary is when you intentionally price something too high so you won't get it but you end up getting it, you get really nervous that you missed something big!

"The first 90% of the project takes 90% of the budget. The last 10% of the project takes the other 90% of the budget." That was the advice we gave clients when I worked as a Project/Program Manager for KPMG.

Practical Engineering has been turning my entire college curriculum in civil engineering into short video form. I’m so proud of the dedication over the years. 😊😊😊

My favorite engineering estimation joke goes something like this. There's a project manager who's projects are always on time and under budget. When someone asks him what his secret is he tells them that he goes to the best three engineers and gets an estimate from each one. The person asking the question then says, "and then you average them?" And the project manager says, "no, I add them up." I'm pretty sure I've never estimated a project correctly.

There's an important aspect of this problem that was missed by this video: If you come in with a realistic, expensive bid, you simply will not win the project, if every other firm is coming in with a lower price. The entire industry is financially incentivized to under-estimate. When I was going through engineering school, we had a series of lectures specifically about the problem of cost overruns in the engineering and construction industry. The simple truth was that, for projects of the same type and scale, firms that came in with higher, more conservative bids, simply did not not win the RFP (request for proposal) bid. They lost the project to the companies that came back with lower prices, EVEN THOUGH all of those companies then went on to experience cost overruns that were GREATER than the original high bid. The data was so tightly correlated, it almost looked fake: the higher the original estimate, the lower the cost of the project in the end, while the lower the estimate, the higher the cost in the end. Like you said, Underestimating ends up costing more than the cost it takes to develop a more accurate estimate.... But all the companies who do, lose out to those with a lower bid. The whole RFP and bidding process that defines the industry is to blame. It incentivizes a race to the bottom in terms of estimate pricing.

I’d be interested in a series where you detail every step of the construction project from planning to execution. What actually is going on when engineers plan a design? How are those plans executed by contractors? Etc…

I have a short story. At my last job we had to estimate the cost of a project and give our plans. My plan cost 3.5 million dollars up front with no annual costs and would take 18 months to implement. The other project had 1 million up front, half a million annual, and would take 3 months to set up. They opted against my plan. The plan they opted for ended up costing 7 million up front, took 48 months to implement, and has annual fees of 2.5 million. Yikes!

There is a way to avoid this problem: inform the bidders that NOT the lowest bid but the next-to-lowest bid will be accepted. Suddenly the bids become more realistic.

As a deaf engineer I appreciate the time taken to get your videos subtitled accurately.

Once I had the chance to bid an engineering project using carefully gathered data from a previous project which was almost identical to the one we were bidding. When we presented our estimates to management they rejected them - saying they were too high and "the customer would never approve of that." They went ahead and cut our numbers by such a huge margin we engineers refused to sign off. Eventually we lost the bid, the feedback from the customer being our numbers were "absurdly low." Live and learn.

This is the most painfully relevant video Grady has put out. I found myself laughing and crying and pausing to shout “YES Somebody Said It!” I’m also an engineer who has wildly underestimated costs based on only conceptual designs and a site walk through…The secret is to write everything down and document every decision.

Speaking of miliary costs. True story. My buddy in the army told me about how everytime they go out for training that the squad uses all the ammo regardless if it's for the actual training. The paperwork to put back ammo is longer form than to take it out. One time they blew up a whole box of grenades because it was easier than to return it. This happens every single time for everyone. Probably half the military budget is just wasted because the paperwork is a pain.

I've worked on projects that had to be rerouted multiple times because we kept finding things during the environmental survey, but it's better than not doing the survey and finding human remains in a front loader. There's so many old, abandoned, and unrecorded cemeteries in the US that sometimes we were the first people to see them in a hundred years.

Wow this was so cathartic for me. I’m a software engineer and we have a huge problem with project estimates. The funny thing is, we tend to say this isn’t construction and therefore it’s much difficult to estimate something that’s never been done before. But it’s reassuring to know that even construction has a similar problem

15:30 This was my only real takeaway in project management class: Don't be fooled by low costs at the start of the project (=the planning process). Make sure you spend a lot of money there.

I’ve written several papers about this so it’s always great so see more people talking about it. I would like to add one thing I’ve talked about a lot, and that is the way construction bids work they favor lower costs and hence underbidding. If you make a budget that covers those extra expenses you will likely get underbid by someone that says we won’t need those.

I’m a retired construction superintendent who managed commercial building projects up to about $12 million (todays dollars about $20 million). The thing that drove me crazy was low bid was the deciding factor. My grandfather and father were also supers and in their day low bid and high bid were dismissed and the middle bidders were then analyzed for ability to perform. Near the end of my time general contractors I worked for would rush out to sign the lowest bidder before they had time to find their mistakes

I’ve really come to enjoy the topics you cover in your videos and how it relates to things I’ve learned working in the commercial construction industry. Keep up the good work

I definitely see a correlation between the fee I have to do drawings and the number of change orders later. It's really easy to move a wall digitally!

I was an engineer on the Big Dig from 97-99. It wasn’t uncommon to see stuff like three guys watching a robotic welder performing operations on the slurry walls. And everyone seemed to charged 2x hours to the project. I remember being amazed that it actually finished. The cost was not suprising 😅

You could compare the cost overruns of large projects undertaken by corporations vs. that of government bodies to see if there is a significant difference (I.e. new office building vs. new city hall)

The other thing worth mentioning is that there's almost always some kind of competitive bidding going on that provides an incentive for parties to underestimate costs. A bidder who bases their estimate on everything going perfectly is obviously going to have a lower bid than one that realistically considers the risks...

As a design professional, I relate to this so much - so many factors go into the budget for a project and there are so many variables that can affect the budget. Even if one can get certain costs with reasonable certainty, time escalation and market conditions will change things with the passage of time. Projects take many years between feasibility and actual construction, original budgets escalate significantly over 5-10 years. The client and/or funder never really wants to see the costs going up and the need for a bigger budget either.

Thanks to you and your videos, I have been able to understand civil engineering to a much deeper level. I have been watching your videos for +5 years, and they are always amazing!

In the UK there is a rail project called HS2 that was originally going to cost £30-£36 Billion and is now costing around £100 Billion. The project started around 2010. The first phase is estimated to be completed in 2029-2033 and the second phase 2035-2040

I have estimated many projects and I was directed to underestimate all projects in order to win the bid. My real estimates were always close. Also, I focused on the time rather than the costs to complete the projects. Time is always more important than price. There is no incentive for an engineer/estimator to over estimate a job and lose the bid. Also, especially government jobs, management/clients cannot keep from meddling in the project inevitably increasing time and money

Thank for your insight. I do appreciate you putting these together. Being involved in a number of large construction projects over my career. What you said here was great. How we budget and bid projects really needs to change. My personal opinion is that. Establish what is the goal, and the life cycle, maintenance, operational characteristics, length of construction, and perhaps a few other parameters. Then put it to contractors to provide bidding based on these parameters. Make them responsible for oversite and unforeseen stuff. Make them responsible maintenance and operation for 10 years. And at the end of that have 3rd party bond release to prove it is on path for life cycle. And bond it for life cycle. You would need to allow for innovation so the construction costs could be mitigated where possible. We proposed a system on a large scale project that would have saved as much as 30% in direct cost on the foundation and cut 6 months from the over all schedule. We were willing to prove the concept at our cost with 3rd party oversite. But the answer was "we have a way of doing things and this doesn't fit". There wasn't an incentive to save money. It makes me suspicious. The savings may have been $40m or more. Wasn't the first or last time I bumped up against that type of mentality. It happened in my own company. So, I retired early as did another bright person I know. Until we change things we should expect things to continue as they are. Again thanks for posting these TRvids I find them informative and it keeps me thinking even if I am not involved. Mike

There's always something a little ridiculous about budgets. Let's say you have a project in mind, and there's a 35% chance it costs $100k, 30% chance it costs $150k, 20% chance it costs $200k, and 15% chance it costs $300k, due to various possible contingencies during the project. If you wanted to provide a single number to the decision makers, you could give them the 'expected' cost of the project by multiplying and adding those together to get an Expected Cost of $165k, but if you've budgeted your contingencies properly, there's 0% chance it will cost $165k! And as Grady said, you can factor in inflation, but since the 2008 crash a lot of countries have implemented good governance regulations that outlaw back-of-the-napkin guesses since they can be used to nefarious effect. Speaking of nefarious, history has had more than a few cases of the underhanded Robert Moses tactic of "Tell them it'll cost 1/3 of what it will actually cost; then once a politician's staked their career on this project, you tell that politician they need to find the rest of the money or else the public will blame the failure of the project on them."

I do plant level project engineering, doing this process for 15 to 20 projects a year. Really good summary of the challenges of it!

Excellent summary! As a civil engineer and project manager, I was involved in cost control in various roles on all sides of the table, and I can relate to these explanations. However, even fellow engineers do not understand why infrastructure projects today apparently always run over the budget, while in the past, that was (allegedly) never the case. And the recommendation around the 15 min mark is really to be considered. Every € you try to save by cutting corners in the planning phase, you will likely pay 10 times more in the construction phase. Of course, in large critical infrastructure projects, you sometimes do not have the time for complete planning but need to start asap and fill the gaps later.

I always appreciate your in-depth videos to better understand how construction is done and what challenges are faced when doing construction jobs 💪 Most of my friends always complain about construction jobs but I always thought there are good reasons for why projects take so long etc. Thank you for your videos!

Excellent video, Grady, probably one of your best to date, and that is high praise. Very topical and relevant - here in Sydney, Australia, we have some major infrastructure projects that have exceeded budget and your video helps to explain why. I also appreciated your comment at 4:15 : "Engineers are just not very good at estimates." The same goes for politicians, public servants, contractors, and just about everyone else, I think! Your suggestions at the end on how to manage expectations and risks better are sound, and should be adopted widely.

You’ve done a great job covering all aspects of cost overruns. I work in building construction and can vouch for the complexity. It’s like planning a car journey from Europe to Australia and trying to design and plan for all the known roads, known towns, border crossings, ferry rides, etc. then you begin and find roadworks, rough seas cancelling boats, civil unrest breaks out, cyclones, etc. The most successful projects I’ve worked on are Early Contractor Involvement / Engagement where there is an early tender process for rates, markups, attributes, and value proposition. From here, a Contractor can add much more meaningful value with realistic budgets, timelines, build complexities, etc. And the myth that it turns out more expensive is bollox. Clients can save thousands in PM and QS fees if the Contractor is doing it. Then I’m construction, the Contractor is usually open-book and has his reputation to uphold in keeping within his own budget.

I worked on designing temporary support structures for the Big Dig. They deliberately under estimated the project so that they could get it through the legislature. Previously, as a town engineer, I designed and ran my own small bridge replacements. I did them for less than 1/3 of what the DOT estimated and only took 3 months to replace the bridges. I started construction the day after the last school bus went over in June and had it open again before school started. Consultants complained that I was denying them work so the DOT banned me from doing bridges. All of a sudden, large bridge projects that I estimated at $500,000, became $3,000,000. I'm now glad that happened because it was just before chinesium started taking over and causing problems. A very good and thorough video. Yeah, the idiots don't understand that construction escalates at a much higher rate than the government estimates inflation. When I was a combat engineer bridge builder we were taught to add 25% for losses, on our estimates. I carried that over into my civilian estimating. I hope that young engineers are watching your videos. Good Luck, Rick

Great video Grady. The scale of the unknowns is truly hard to comprehend at the beginning and hard to explain why they weren't considered at the end. Another point I would add is, often estimate are right and projects do finish on time and on budget, but since what happened is what was supposed to happen, it's unremarkable (and not news worthy). People only remember the extremes, Thanks for another great video!

I've always been impressed by the way you break down complicated topics and provide just the right degree of detail with am eye to what's practical. So I suppose I shouldn't have been surprised to see this topic. But my first thought was "Oh man, how is he going to cover something this huge". Amd honestly, you nailed it! Thanks for another great video!

A lot of large projects I’ve been apart of as an electrical estimator come down to how rushed the design and tender process is. Often we look at the plans they give us to quote and there is so much missing/required on top of what is presented to us that we just quote what they ask for knowing full well that even if we quoted close to cost price we are going to easily make 10%-20% once variations start being applied.

I always love watching Grand Designs and taking bets with whoever is watching with me on how much longer it's going to take and how much over budget it's going to be. As a general rule of thumb 1.5x the Time & Money is a good result, most is 2x and I've seen it go up to 3-4x on other projects.

I converted to a Henson when the cost of cartridges got ridiculous. It took a bit of getting used to, but now I wouldn’t shave with anything else. It does a really good job for a tenth of the cost. I get ten decent shaves out of every blade. It is easy to keep clean and to swap blades. My kit came with lots of blades, a disposal container, shaving soap, alum for nicks, and after shave cream. I already had a brush, so swapped for more shaving soap. Set for life.

Grady, thanks for putting this issue in perspective. I once managed an environmental remediation project which I estimated to be $850k. At about the $700k figure I warned the client that the final cost would probably break $1m based on field work to-date. The final cost of the project ended up ~$1.2m. Thankfully, my client had the ability to fund the project to completion. I, however, felt very bad that my initial estimate was so inadequate.

I work in software engineering, and I am not surprised at all. We do mainly in-house development within a medium size corporation. Most projects take more person hours, and also more actual "calendar" time than initially estimated. I think the main reason is that many complexities only truly unfold when you actually do the work. Also, more often than not, our customers wish more functionality than originally planned, so the final product is better than expected, but also more complex.

Every project is unique, there are so many variables it is mind boggling. Changes are inevitable, unless the documents are perfect, and there are no surprises on site. As a contractor I hate change orders. They delay the project and tie up resources. The best jobs are the ones where everyone works together to make sure critical milestones are met and problems are sorted out quickly. Unfortunately this is not usually the case. A cooperative team that includes the client and the constants can make the difference between a good project and a disaster.

I do agree with most of the points. I would also say that these extra costs are not limited to construction projects. In Oil&Gas, chemicals, any other projects, there are always budgeting issues. Adding to all the elements mentioned, there is one particular thing that can turn projects into disasters: extra features, or modifications added by the end customer. We have a running joke in Europe about The A400M, the joke says: do you want to know what happens if you let military decide about their plane ? Look at the A400M...

I think almost every project I've worked on in the last 2 years was budgeted 3-4 years before their start. As a result all of them were over budget before engineering even started. I also end up having a lot of clients who want to change or add things to the project which add cost and then are amazed that the new estimate is higher than the number they had pre-enginerring. Then they ask you to do anything you can to cut costs putting me in an awkward spot where I have to figure out which of my standards I can skimp on without actually compromising the safety or functionality of a project. I think this is why there are so many awkwardly designed things in the world that I question.

This is so interesting. There should be a meta-analysis to get an estimate of by how much usually construction projects costs are underestimated.

I work in Architecture. Every single architect I've ever worked for has told me the exact same thing. "Never ask an Architect for how much something will cost to build, because we're always wrong."

I watched a very similar video from a non-engineer perspective. Very interesting to see the difference. They were discussing how humans are bad at estimates and you talked about the challenges such a task entails; including everything that could and will go wrong.

Even without all the materials cost issues in civil engineering we had a similar issue in software development. We have the 90/90 rule. The 1st 90% of the project takes the 1st 90% of the time, the last 10% takes the other 90% of the time.

As a former Bostonian, I'd say that the Big Dig has had an very positive impact on the city.

Agreed. Do all of the engineering possible up front, then when pricing it out, calculate risks. After so many thousands of projects have been done, surely there are some numbers that can be used to estimate the likelihood of delays, problems, redesigns, etc. Some projects are probably much simpler, using well tested methods. Innovative new methods have higher risks. Then when presenting the price to the public, it's "if everything goes as planned then the price is ____. But it's likely there will be issues along the way do the price is likely to be higher." Then offer a more likely estimate.

Great stuff! My own experience also notes: not only do the higher, more realistic bids not win awards, but also, nearly every company’s management is trying to get you to work the price down, not up, before it even goes out the door to the client. (And God forbid somebody gets the Sales folks involved!) If you dare to identify a budget contingency or schedule contingency as a percent, management will generally dictate a reduction to the percentage, with no in depth review of how the number was determined. And finally, schedule: in any network of dependent tasks, schedule gains are rarely passed on, but delays almost always are. Critical path/PERTT planning pretty much locks this fact in. There are other scheduling methods but they are not often used for big infrastructure projects. And thus, when you almost inevitably overrun schedule, you’ll very likely overrun cost, because many overhead roles (including the project manager) have a cost essentially dependent on days elapsed.

I learned this lesson the hard way when building a bed frame for a customer, did everything I could to properly estimate the cost including calling the lumber yard and getting the most up to date price. I get the customer to confirm the project about a week later go to pick up the lumber and now the lumber prices are up by a third and the "saftey" money I had calculated into the project wasn't even enough to cover the diffrence. I can't imagine trying to plan 10 year infrasturcture project because there it literally no way to know what something is going to cost a week from now, let alone 10 year!

I just thought I'd mention that this kind of thing can also happen with software engineering, except it's usually over schedule rather than over budget. Feature creep and bug fixes can spiral out of control and a project can be months if not years overdue. Engineers just generally aren't the best at these kinds of things lmao

What hurts the most is although the estimate is wrong, I’m responsible for the overage as the civil and structural construction manager and that causes more stress on the project than you can ever image. There’s a trickle down effect into the core craft group that can lead to loss of qualified tradesmen on site. The stakeholders don’t care why, they just care that it went over and they need to find the extra money, and it can snowball when stakeholders try to hold the budget down when the professionals are screaming for more trying to get it done.

I was in the military and was in charge of overseeing construction of a new building. Several of the contractors didn’t read the contract info carefully enough and completely painted themselves into a corner. I had the sad privilege of getting to inform them that they’d done the work all wrong and had to redo it.

Great video and very well explained. As a project director this really resonates with me. Btw I'd be getting a quantity surveyor across those estimates especially if an engineer has prepared them 😉

The trap I fall into, and that I'm always trying to get better about not falling into, is estimating based on everything going perfectly. Things almost never go perfectly. Now a days when someone is planning a budget and asks me how long something will take I take the first figure that pops into my head and almost double it and I often end up being much, much closer to reality.

As a project engineer in Solar Rooftop business where most projects are just cookie cutter of another one and there's a lot less complexity compared to construction. My team still runs into overbudgeting issue as no roof is the same, and clients are very diverse. And the bean counters wants us to underestimate as lower cost usually gets the contract. It's a never ending cycle.

There are tons of reasons, but they tend to boil down to 3 categories. The more unique a project is, the harder it is to effectively estimate. If you build the exact same power plant many times by the time you get to #3 or 4 you have a great grasp on what is going to happen. But unique projects have many of 'things you think you know, but are wrong' and 'things you don't know that you don't know'. The less experience the people involved have, the worse the estimate. This frequently compounds with #2 but can be independent when you get a critical mass of morons, nepotism recipients, or consultants in charge. And if it is public money there is little reason to come in under budget and plenty of reasons to pad costs to benefit unions, organized crime, political machines, and general graft.

On my current job my company had to add an exra $15,000 charge because the ground turned to solid sandstone three feet deep. Personally I think it should have been redesigned because the rock was as hard as the concrete we were replacing it with. But I guess its easier for them to just add to the budget 😅

I'm an DOT engineer. We do have data on material cost, number of projects under specific construction contractors, and grading in the recent year. 80 - 90% of the construction projects are about 15-20% more than estimated. We are using all these data to help our designer/consultant to make better estimate. An experience engineer can tell if the cost is way greater than the overall efficient. These political projects are the real problem. Not only we are giving a limited time for the design and construction but these projects almost alway go over budges.

Great video! Very insightful for entry level project engineers. Keep up the good work 🙂

Excellent video. The reality is that even if you can perfectly plan your project, base your estimate off of 100% design level and resovle all the issues covered in the video, construction is still a commodity. You won't be able to accurately predict the cost of a project 4 years from now in the same way you won't be able to predict the cost of other commodities like oil or steel. And since the incentive at every level is to underpredict, the costs will end high.

Hey! Thanks for not forgetting about surveyors! Everyone else seems to lol I have literally had contractors ask for a cheaper price because they "forgot to carry survey costs in their bid". This was a multi-billion dollar building construction!!

Bids should include a history of the last say 10 projects and the % they ran over. The that can be used as a criteria with the bid. This goes for engineers and contractors. I started using $/lb for similar work that was done for estimates and they always sound so high but then aren’t. If course that takes a library of estimates to do but it beats the old guess and check method.

Love the videos Grady. Keep them coming. I think you hit the nail on the head pretty early in this one. Enginerds. I'm an electrician. Almost 30 years in the IBEW now. And things are getting worse. Last job and current project have had beautiful 3D CAD models created by enginerds. OK, expecting the computer equipment to see them to survive in the field is the first fail, but... Seriously? Go start wrapping a rope around a tree. 1/4 turn, you can pull it. 1/2 turn? Yep. 3/4? OK, but getting harder now. Full wrap? Umm maybe? NFPA NEC has a limit of 360° of bend in conduit run for a good and practical reason. And yet the enginerds are producing the gospel 3d model for the job with 5 90° bends, 2 offsets and a kick. With no additional pulling points. In a run that is too long to be pulled. Oh, and the current project is in an EMP shielded enclosure, so adding supports for an additional junction box requires congressional approval and the blessing of the Pope himself before we get into the cost of the additional materials and labor. Why do I only get paid labor wages to do the labor and re-engineer the failed designs? Pretty easy to see where the cost overruns come into play when the designs aren't up to code or reality to begin with.

1. Good Show! I like your productions more every time I watch one. This one I even enjoyed the "commercial". 2. I was surprised in my software career by how much my time estimates were relied on; my rule of thumb was to take my gut feel, considering every thing I could think of, then add 150%. I figured I'd rather endure the shouting at the start, and occasionally enjoyed the kudos when it came in early. But even given that rule, too often I could only save my timeline by putting in more time than I thought.

I'm studying a short professional course in project management, and this video put a big smile on my face. This is beautiful.

We have one of the largest civil engineering earth-moving projects in the western hemisphere located here in eastern Kentucky. The Pikeville Cut-Through moved 18,000,000 lbs of soil (the Big Dig moved 15m), rerouted the major highway, rerouted the railroad tracks, and moved the route of the river. Additionally, it created 390 acres of usable land for downtown development. It was completed in 1987 at a cost of $77 million.

I recall that on RM transit, they pointed out that the rapid expansion of high speed rail in Spain was on or under budget. The reason was the companies were experienced in such projects. They had very few 'back to the drawing board' moments. Stations were similar, and not excessive. Unlike the big dig where there (fortunately) was only one instance, and everything was new and different. As an aside, I wonder what 15B would have bought in regional transit improvements.

I really like the objective practical approach to this video. You’re statement about a lack of practical experience could be applied to a great many people who are well intended but highly inexperienced. Thank you!

My job is to deliver large CapEx projects on time and with budget in mind. That being said, there are tools to estimate the cost to an acceptable level. In my opinion the problem is that companies start their projects without a project management and engineering team of professionals.

Maybe reporting two budgets would be useful: (1) as designed and (2) anticipated final design. The anticipated final design does not mean guess the actual design that will be final, but rather, estimate the overrun that is likely for such a project given that change orders will come up and cost overruns that are typical. This is a statistical prediction problem -- there could even be prediction intervals on the second number to express the level of uncertainty (which is probably high). That said, I wonder if a new challenge would arise: if we have a proper, unbiased final estimated cost, the bigger $ number could attract even more fraudsters, driving up the cost even more.

So I'm an industrial carpenter and I have seen some stuff in the 10 years I've worked in that field. The other day I was asked to install some regular out of the box television mounts on some specialty surface walls. Seems easy enough and they gave my partner and I two days to complete it. Now the curve ball. My company decided that because it was going on a specialty surface, that they wanted to send it out to be engineered. So now as a good little field worker, I need to install it as it's been re-engineered. But they bought me all the wrong hardware and the wrong materials. Luckily other trades loaned us the hardware we needed and another scope of work had the materials we needed. But it took us 2 man-days to fix the hardware and materials mess up from my office, re-fabricate our new supports and get ready to install them. All still inside the 2 day window we were given. We wound up over budget obviously and received flack from the customer because "how hard can it be to just install some TV brackets on the wall?"

There's a fair amount written about how the Hoover Dam came in 2 years early and under budget (despite being the largest dam ever built at that time that required new construction techniques) that makes for interesting reading on this topic. For example, it was awarded to the lowest bidder, but the difference between the two lowest bids was due almost entirely to a single line item (estimated cost of concrete) and the bidder's cost was only $24K higher than the government's estimate (out of a $49M bid cost)! The key incentive in the contract was a $3K/day penalty for being late on any of the 5 portions of the project. The six companies involved in the bid had collectively completed over $400M of projects prior to bidding on the Hoover Dam, so had a wealth of experience to bring to bear, and they picked the right man to lead the project.

One (small) job I did an estimate for, got questioned by another engineer (not unreasonable). However, in discussing how I got the person-hours to do the work. I started explaining that you start from 2080 hours, subtract holidays, annual leave, a week of pto/sick leave, hour for weekly meetings, time for mandatory training, and at the end you might get 1800 usable hours in a year. He said, oops I never thought of all those “lost” hours in my estimates, time to account for them in the future. So always room to improve estimates.

I worked on a US Made Only public works project where most of my time was spent drawing and detailing and communicating and getting approved orders for basic parts like bolts and screws to get them made by American machine shops simply because German or swedish products just weren't good enough. Made the cost of the project significantly higher than they projected because they didn't realize the scale of a lack of us products. Some of the things we just substituted foreign materials without them knowing simply because we wanted to move on from the project

If all projects were linear and timely, extra costs would be easier to cover with planned contingencies. However projects get bogged down in permits and approvals ( especially with public projects ) that don't come with a predictable timeline. As we know, the price of things never goes down with time....

This super informational video is yet another reason why this remains my favorite TRvid channel. You explain a complex topic so concisely and make it very interesting. 👍

Different in detail, but the same in principle, this is a key aspect of software engineering. It’s accepted that ‘wicked problems’ are the norm, and have refined methodologies to handle them. I observe that many industries are learning from these, and many more should. It’s simply reality that we can’t know everything up front and 100% predict the time and effort complex projects will require. But we can improve our estimates for sure.

Small correction at 4:25: as an engineer often tasked with estimation during different phases of the design, we absolutely do track material prices, labour cost, .... You can't estimate the cost of a quay wall without somewhat recent steel and concrete prices. I'm assuming other countries have indexes for those as well.

Thank you from the bottom of my heart for this reasonable and rational based video!

I'm an ironworker, and I will agree any day of the week to spend more earlier in the process. Mistakes do nothing but get bigger and more inconvenient the longer they lie.

I've been looking at Henson razors for a week or two now and was going to get them earlier today. I decided to wait until tomorrow, and I'm glad I did. I think I would ultimately have gone with Henson regardless, but now I get an extra 100 pack for free and I get the recommendation of you to go along with it.

At Seatac Airport they had a Sky Bridge estimated at 375 million, go over budget by 700 millions dollars. I would love to see a video on this.

Excellent video. This aspect of construction is not shared enough with those outside of the industry. Keep up the good work.

You can make any construction project come in under budget, if the budget is big enough.

This is an excellent summary. I come from the contracting side, and while I was never involved with bids, I was involved with 'finding' extra costs during construction. In theory the contractor gets paid whatever he bid, but of course that depends on the type of contract. For larger civils jobs there is always going to be a degree of uncertainty, so there is an agreed method of paying for extra, unforeseen, work. But there are also fixed price contracts for smaller works. In this case the theory is that the contractor takes the hit for unforeseen costs. But of course there are always means to be creative. Go back to the drawings and schedules to find something not clearly specified and give the client the option of paying for something substandard he hadn't anticipated, but what you have priced for. Or give him the option of paying for what he was expecting, but adding a nice markup. Creativity isn't always that hard. One of my first jobs was on a site for the UK ministry of defence - a submarine base. We were building jetties with a few brick structures and service roads. The detailed drawings showed they had specified a certain brick manufacturer and brick type and finish. We had priced for this but when we told them we had only priced to cut standard bricks to the angles needed, they demanded special bricks be made for e.g. 45 degree corners. As we hadn't priced for specials I prepared a huge schedule of specials for the manufacturer. Of course the MoD had specified one manufacturer, so they added a huge markup, as did we. This was probably a small gain, but multiplied over several other claims, turned a high turnover, low profit job into a healthy one for us! Others have pointed to low bids as being a problem. Yes, there are contractors who will bid low in the hope of claiming extras, but it is a fraught time-consuming process, not something any contractor wants to do if they could get a reasonable return without it. Saying clients should accept higher, more realistic bids is easier said than done. Those higher bids are probably from contractors who already have enough work in the pipeline. They bid to stay in the game for future contracts, but know that others will bid lower, so don't expect to get the work.

Just out of engineering school I was given a cost estimate project by my boss. A section of a steep hill/ cliff cut was to be set back and anchored where necessary. It was a fairly simple project but the unknows of the rock type and the stipulation that the project owner decided on the type of construction to be used as the project progressed made it impossible for me to make an accurate estimate for the project. I was able to make estimates of cost per foot for each of the types of construction that could be called for by the owner but not for the entire project. We didn't get that job but, my boss told me the company that did had a person whose sole purpose was to fille change requests with the owner to increase revenue.

It'd be interesting if instead of a cost for a project, there was a range cost. $X for if the project goes perfectly, and $Y for expected scenarios. Although even this will undoubtedly still have many projects going over even the $Y estimate (archeological discoveries for instance) but would give a clearer picture of the variability that may exist.

I work for a large engineering company managing clients budgets, costs and schedules. One of the largest contributing factors is cost escalation. Estimates can be carried for years during planning and when final design is being done, there can be tremendous pressure to "maintain the budget". Estimators need to speak up and tell the client what they need to hear. Also, the Big Dig was almost entirely underground so I'm guessing that differing/unknown site conditions created even more schedule uncertainty. Over the 30 plus years of planning, design permitting and construction technology has changed a lot. 10 years of escalation at 3.0% annually would add over 34%. We are doing programs longer than this where escalation is over 50% and is by far the largest risk on the risk register. I'm not even going to bring up scope creep where clients add scope to the project. With the recent escalation we've experienced that has become less of a problem but I'm sure it will come back to us as prices normalize.

When I used to have to try to figure out how long a project might take, we had a saying: The first 90% of a project takes 90% of the time... and so does the last 10%! Not that accurate time wise but it conveys a key idea that it takes more time than you expect.

There are a lot of incentives to under-estimate and not a lot to over-estimate. There is also the fact that it is easy to forget about something that adds costs but you rarely forget about something that lowers the cost.

What a fantastic overview of project costs and the uncertainties within. I work in excavation on highway jobs and this will be my go to video whenever anyone asks why did this road or that project take so much longer or cost so much more?

I think that if a lot of these large-scale infrastructure projects had a truly accurate cost estimate up front, then they just would never get built. Sometimes optimism on the front end ends up being a good thing. Other times, you get the Super Collider.

I was living at the original Fermilab long ago and remember the story that it came in well under budget. It may have also been ahead of schedule. This was because they had an actual scientist in charge, not someone with an ivory tower degree in who-knows-what.

I'm a senior project scheduler, having worked in the heavy construction industry (think power plants, etc) for over 2 decades. Cost & time estimates are often more art than science. I keep a sign in my office "Remember Cheops' Law" (Nothing ever gets built on schedule or within budget - thank you Heinlein.)