project time and cost estimation

[Admin]

One thing I've always struggled to do is estimate the final cost and time it takes to finish a project. Typically I'm off by a factor of 2x or 3x, and always an underestimation . . . I always forget something, like time lost while waiting for stuff to ship, or taxes, or factoring in time for dumb mistakes.

So in an attempt to fix this problem of mine, I wrote up an excel sheet that reminds me of all the things I should take into account, and adds it all up for me.

http://www.societyofrobots.com/downloads/SoR_project_estimator.xls

Let me know if you find it useful, accurate, or have ideas to improve it. I'll put it in the Misc section on the main site in a week or so.

Also, for those experienced out there, have any recommendations to improve project estimation accuracy?

[jman571]

Hmm, a lot of that stuff doesn't apply to people who are making just orders for small parts (as in not pre-fabricating their own circuit board  ) but I completely know what you mean Admin.

In my opinion it's always the small stuff that you don't bother to factor in, that adds up in the end. I mean it ended up costing me $25 for my custom-cut acrylic sheets, $8 for the chemical bond glue, and even things like purchasing a cheap "shopping cart wheel" caster, various bolts, etc. added up to unexpected costs. I think it's all about what you start with though. In my case, just about the only thing I had in advance was some Velcro adhesives, which were incredibly useful in securing my battery and my receiver, but if I didn't have any Velcro I probably would've bought some anyways.  So for someone like me I needed to buy EVERYTHING  If it wasn't for my uncle I wouldn't even have any threaded rods, nuts, or even a power drill  Yes, some families do not even own a drill   ...I guess my dad was never much into DIY, which is funny because for work he built stuff at home for a while, just never needed a drill I guess 

EDIT: One more thing I ought to mention, your excel file brought to mind an old quote: "Time is money". I think if you have one category filled out with something like 50 or even 100 hours spent on (in my case...this would be research...yah 100 hours  ) and yet $0, you should consider if there's a faster way to do it, but you may need to spend some money on it. Now this does not apply universally, as somethings do need to have time spent on them, and no amount of money will make the process any quicker, but there is an interesting correlation between time and money, and perhaps in future projects this concept needs to be given closer scrutiny. This has some far-reaching implications, but I have to finish my robot project write-up for tomorrow , so will not be able to expand on this idea tonight 

[S. Karim]

In the first totals, why are B24 and C24 (misc.) not included into the math? For B25 for example you have =SUM(B14:B23) and the Misc. Hours is the B24 cell, so it's not adding it up.

Same for the other totals?

[Admin]

oops . . . ok its fixed now.

I had added the misc stuff as a last minute thought . . .

[SixRingz]

. Typically I'm off by a factor of 2x or 3x, and always an underestimation . . . I always forget something, like time lost while waiting for stuff to ship, or taxes, or factoring in time for dumb mistakes.

Admin, there's a universal rule that explains this and will hopefully help you do better estimates in the future. It's called pi-factor and is applicable to alot of things. Money and time are both subjects to this. The formula is easy:

Outcome = estimate * pi

Once you start using this, you will be shocked how great it works. 

[4by4]

The 2x and 3x factor is so true. I saw that over and over when I worked on big government projects. We had a rule of thumb to always double our estimates, but then our bosses knew we were padding, and would always cut back the estimates. So then the dilemma was, well, shall I quadruple the estimate, because I know my boss is going to cut it in half?

[bukowski]

Hmm. So does

(Outcome) * pi = (Estimate * pi) * pi   ??

[SixRingz]

Hmm. So does

(Outcome) * pi = (Estimate * pi) * pi   ??

Haha, what is that supposed to mean?
Anyway, doing estimates is hard. My suggestion is to have a template for all projects and go through that every time you need to estimate and then also when the project is finished. That way you get a good feeling about the fields with large errors and also you might find re-occuring things over projects in a "Other costs" post or similar which can be added to a revised template.

[bukowski]

dumb circular logic joke. sorry. 

[izaktj]

Hmm. So does

(Outcome) * pi = (Estimate * pi) * pi   ??

lol so the pi on the left cancels the outer pi on the right haha

[Admin]

go through that every time you need to estimate and then also when the project is finished.

Good idea! This is why everyone should report back in detail how many hours they spent on their robots. I've only kept poor records in the past . . .