Exponential Bidding on Projects

Ah, but then the cost-minimizing function suggests very short builds. And those are expensive. Better might be a polynomial function with a certain minima at the preferred project length for that company.

The thing about linear is, it leaves no preference, which I guess might be considered the company setting its own pricing terms -- which allows it to maximize profits within the constraint of the overall competitive system. You have interesting points, but they may just not be interesting enough?

Tim

-- Deep Fryer: A very philosophical monk. Website @

If the customer response is...

Arrrrgh! : Bid was too high Arrgh, OK : Bid was just right OK! : Bid was too low ;-)

...Jim Thompson

I remember when pricing some hybrids from an Italian company, the price was fairly steady until we reached a quantity in the low tens of thousands, then all of a sudden the price jumped dramatically. We asked why, and the answer was "if you want them by the deadline, we'll have to build another factory!"

If you have no shortage of work, and you can force a definition of the requirements to your customers, then piece-work pricing makes perfect sense and keeps you from getting involved in the "bring me a stone" neverending projects. Most consulting companies bring in money by doing the exact oppositte, and in fact some customers cannot deal with anything except ill-defined never-ending projects. Don't go there!

Tim.

Ill-defined, never-ending projects are fine as long as I'm getting paid by the hour.

Classic reference, not for consulting but for in-house software jobs, is "The Mythical Man-Month": "Adding manpower to a late project makes it later". You're probably selling yourself as a single consultant but even there I think you're realizing costs related to a "big job" that aren't easily scaled, something that Brooks can neglect.

Yourdon's _Death March_ also talks about personal costs of colossal projects.

Tim.

Maybe from a purely bean-counter perspective you're right... but how many of your never-ending projects have gone on for 5 years? 10 years? There's a real personal and professional cost of getting sucked into those. Many will assume that if you are in such long-term projects, that it's only because you didn't have the wisdom to say "no" - it's the sort of thing that I at least leave off my resume or CV!

Tim.

Sounds like my salary negotiating skills.

Him: How much do you want? Me: 20% over my current pay. Him: OK. Me (to myself): (Dammit!)

Yeah - quote a job that way and see how many people beat a path to your door. >:->

Good Luck! Rich

And if they go into cardiac arrest, you've obviously confused your current customer with that other government contract sitting back on your desk! :)

I read somewhere that the longer it takes for something to happen, the less likely it will ever happen. (i.e, making a decision). If that's true, projects that run over should defintely have stiff penalties, like mountains of cash for the vendor (who cannot be blamed for the Purchaser's inability to get its act together.)

-mpm

No matter what formula I use to bid on a project, I still have to be competitive to win bids. But on occasion, I don't have much competition and I get to play with numbers more..

Take doubling bids for example.. (Starting off very small to illustrate a point.)

$1 to $2...the diff doesn't even buy a coffee. In that area, I can double and it could be deemed insignificant.

$10 to $20.00... the diff just buys dinner. In that area, I can double it and it could be deemed insignificant.

$100 to $200.00... the diff is nothing compared to someone used to paying $1000.00's on many things.

$1000 to $2000.00 .. the diff is nothing compared to someone used to paying $10 000's on many things.

$10 000 to $20 000... if they can afford $10 000 then $20 000 maybe affordable to...

Something like that...

D from BC British Columbia Canada

Yup.. It's just like that.

My thoughts during bidding are either..

Damn!... I lost the bid...I was too high. Vacation time! :) Or... Damn!,,, I won the bid.. I was too low. Vacation cancelled.:(

D from BC British Columbia Canada

Nobody is going to know I used a non-linear function to make up prices.. Customers just see the final bid number.

Example: Day 1: $500.00 Day 2: $510.00 Day 3: $520.00 Day 4: $550.00 Day 5: $600.00 Final Day: $700.00

Total Bid: $3380.00 (Maybe it'll still beat competing bids. )

I thought of the idea of non-linear pricing to make up for the accumulation of unpredictable events in long term projects.

D from BC British Columbia Canada

First, I almost never bid fixed price. Rarely do I have a prospect that is willing to spend the time to specify things well enough that I will do so.

Second, I seem to spend a fixed amount of unbillable time prior to getting an order, no matter how long the actual work seems to take. So if anything, I'll give a discount to longer-term projects.

Third, there are no "surprises". There are just poorly read or poorly written specifications, there are specifications changes, and there are schedule miscalculations. If I can't read a specification that's my fault. If I can't identify a poorly written specification it's 'cause I can't read the spec, and it's my fault. And if I let a specification change slide by me without telling the customer he's going beyond spec

-- that's my fault.

That's interesting Tim,

I hate by the hour jobs and try to get fixed price including materials (boards, components, big bought in chunks like computers and test gear). I use the same rate regardless of job length except for very short ones where I charge more, at least in part to cover the unbillable starts and finishes. I like to be running several jobs at once - then when things get delayed on one I can pick up an other. Fo me the ideal job is 8 weeks work with 16 weeks to do it in, not too much material. 30% on start, 40% in 2 stage payments and 30% when its done.

Michael Kellett

Most of my calls are something like "The design is broken, we need it fixed", or "We need you to (perform some open-ended research) to prevent a problem". Neither is amenable to fixed price contracts.

This is all just as well, because in spite of my hard-ass attitude I'm not always the best at estimating a fixed price, unless it's something I've done before a few times.

If you are expecting a payment at a future date, you can discount it by your own minimum acceptable rate of return (MARR). That would be something well over the expected inflation rate, of course. If a project is risky you can adjust the MARR upward to compensate. Then decide what rate you're willing to work for at some instant in time, and make a cash flow diagram for the project, then adjust all inflows and outflows to the same point in time, using the MARR and the time. There are standard formulas for present value, future value, and value of a series of payments, all available in Excel. They are based on multiplying a payment or series of payments by the factor [money * (1+MARR)^(time)], with time having the appropriate sign. With that framework, you can model your time as a series of negative payments by you, based on hours and hourly rate, feeding into the overall present value (if you use the present as the reference point - it doesn't matter as long as everything is referenced to the same time). The foregoing accommodates your points 1 and 2. Point 3 can be brought in as opportunity cost - a subtraction from the project's value of the lost return from another possible project, calculated on the same basis. If the result is less than zero, choose the other project.

There's a Wikipedia article on MARR. It's not very informative, but it has some links, and there's a lot of other stuff on the web under "engineering economic analysis".

Nature is not only non-linear but chaotic. You can't predict the weather, so you should use a chaotic function so you can't even predict what you'll be paid.

You seem to think nature can be modeled, so your pay looks like the old hockey stick. :)

Your other points are all invalid because linear pay would compensate you for linear oportunity costs.