Fog Creek Software
Discussion Board

Construction Mgt

So, I'm getting bids for contractors to do a remodeling job - putting a bathroom upstairs.  The three styles I've run into:

(A) Contract for exactly what work is to be done; I have to approve changes as line-items for an additional charge.  Guarentee that the price will not run over (Contactor assumes risk.)

(B) "Estimates" that are guarenteed to be +-30% or so, contract is less precise - Some degree of Shared Risk.

(C) No realistic estimate, bill based on time and materials.  Buyer assumes risk.

Talking to a "C" contractor yesterday, I was suddenly thought "This is like Extreme Programming!"  I'll explain.

The "A" Contractors build a history of estimated price Vs. actual price.  (Say 2 to 1)  They add in a hefy markup for profit (say 2 to 1) and they add in additional "insurance" for risk (say 1.5 to 1).

So the "A" Contractor estimates the job at $1,000 and decideds to contract to doing the work for $6,000.  ($1,000 * 2 * 2 * 1.5)

The "B" Contractor does the guru method, but forces his sub-contractors to promise firm quotes.  That way, if he runs over, he's only talking about his labor, and he's got some room for error.

The "C" Contractor would be best off to use extreme programming.  Divide the project into deliverables ...

For Instance:
    - Tub, Toilet, Sink, Plumbing, Electrical (Min.)
    - 2nd Sink on other side of wall (For Dressing Room)
    - Floor
    - Paint
    - Drywall
    - Trim
Give estimates of these in some form of unit.  As the project progresses, if we start to run low on money, we cut the deliverables at the end.

To make his estimates more accurate, the "C" contractor would be better off getting a history of estimated price Vs. actual price like above.

Of course, some people just don't like risk, and prefer to pass it along.  :-)

Any comments?  I think there's something there.  In theory, this should actually be EASIER to measure and track than software dev because so many of the projects are so similar.  Of course, we'd have to take parts out of the equation and only count labor.


Matt H.
Tuesday, June 18, 2002

The entire construction industry is based on risk and reward.  You hit the nail on the head with your three options, which you see on a much larger scale with General Contractors, Construction Managers, and the guy that does both - Design Build. 

The real key in construction is understanding a few  concepts (same in most businesses).  The first is we are all in this to make a profit.  If, as the owner you try and screw the contractor you will get what you pay for.  Relationships are key to picking which guy you want. I tell my people all the time we are a service company that happens to be in the construction business.  Owners that want the low price or are willing to not allow you to make a profit are not worth chasing.  The wise owner is the one who get three numbers, if one is way low, he asks the low guy to double check his bid instead of jumping on it.

If you can find a trustworthy guy the B or C guy is your best bet.  I will now tell you that 90% of the time owners go with option A.  They get a number of competitive bids, to keep the bidders honest, then get the hard number that allows them to establish their budget.  A poorly defined scope of work will also lead you into a trap with A. (sorry Mr. Owner that is a change order not in my scope of work)

C is your best option to get exactly what you want with some regard to controlling the costs but understanding everyone needs to make a profit.  B is a compromise between A and C.

John Knowles
Tuesday, June 18, 2002

Contracts in internal IT shops tend to be either A or C types, leaning towards A.  Then it's renewal time.  No such thing as an airtight contract to put all the risk on the vendor, but it doesn't seem to stop us making them.

Internal IT shops don't normally operate under a profit basis either, and are usually considered cost sinks.  I think that is unfortunate and develops a lot of lazy minds.  Nothing like competition to keep you sharp.

Joe AA.
Tuesday, June 18, 2002

Ok, John, you have the MBA ... why isn't "A" the best bet for a business owner?  (Contractor?)

Here's what I'm thinking:

  Under "A", you get to put aside profit + risk.

  If the project runs over, you eat into risk.  If it runs way over, you eat into profit.

  HOWEVER, over a series of projects, your costs tend to normalize - some run over, some run under, and things average out.  If you tend to estimate right, you tend to have an account for Risk that gets larger and larger until you can skim from it as profit.  If you tend to estimate slightly too cheap, then you still get your profits, and your risk ("escrow") account still tends to balance itself out at $0.00.

  This requires rather descriptive contracts and some good-faith, but I think it works.  Over time, you tend to be compensated for the risk you take.  Of course, you could say that over time, the homeowner gets the best deal with "C" (I think you did say so) because the home-owner absorbs the risk and doesn't have to pay the extra mark-up for it.

  hmm.  Leads me to Heusser's Theorm:

-> Instinctively, Nobody Likes Risk.

That's why I like the XP ideal of shared uncertainty.  It sure beats the morale result of trying to off-load things unto the other guy. :-) 



Matt H.
Tuesday, June 18, 2002

The down side to A is risk-reward works both ways.  I am Mr. A, I bid it as this much, using this much stuff with this much time.  I found a different way to do it (that outside the box crap we keep hearing about) which will save you a ton of money but it is not worth it to me to sit down and talk about it.  Hey if I don’t do a part of my fixed contract you will be asking me for some of my “bid dollars” back.  If you structure your contract right you put in the ability for the contractor to do value engineering and you give him a part of the savings.

John Knowles
Tuesday, June 18, 2002

There is one other thing to remember, size (more $$$) and complicated projects can dramatically affect bids. 

We make really good margins are certain types of work. Low cost ($2,000 - $8,000).  Easy to bid, low in labor hours, we put in a 20-25% margin but if things go well I have seen margins in the 60-80% range.  Now and then we get a problem child and break even,  we almost never lose on this type of work .  The problem is when you try and estimate a large complicated job if things go bad, they really go bad.  The flip side is everyone wants that big dollar contract so you better have a sharp pencil. 

Matt, your right nobody wants the risk.  My old company’s bread and butter was risky jobs.  They did very well at it and even when they had the occasional loser, their margins on work was 4 times the risk averse company’s.

John Knowles
Tuesday, June 18, 2002

For me the real problem with type A is it creates a confrontational relationship with the customer. They want to get every last drop of blood and you want to keep costs down to maximise profit. I am especially wary when the customer is demanding a type A job.

Tony E
Tuesday, June 18, 2002

I can relate to this, as I'm in the middle of talking to builders about renovating my house.

With type A, you have to specify everything you want in the contract. It's hard to change your mind part-way through, and anytime you want to do so the builder gets an opportunity to gouge you for more money.

With type C, you better trust your builder. If you're paying him by the hour, you want to know that he's actually doing stuff, not just looking busy. You want to see progress. But at least you retain some degree of control over the project, and can add or remove things along the way.

I think we'll end up going with type C.

Darren Collins
Tuesday, June 18, 2002

"""I think we'll end up going with type C. """

You are a very brave man.  If I knew what I wanted, and the contractor wasn't able to estimate, I would assume that he is incompetent and/or inexperienced.

Johnny Simmson
Sunday, June 23, 2002

*  Recent Topics

*  Fog Creek Home