estimating projects

Steve Meyers steve-plug at
Tue Jan 10 11:37:23 MST 2012

On 1/10/12 10:09 AM, Nicholas Leippe wrote:
> One of my professors simply said, "If it's a hardware project, make
> your best guess then multiply by pi... If it's a software project,
> make your best guess, then multiply by pi squared."

At a previous employer (where you and I were co-workers), I was asked to 
give a quick estimate on a project.  We had a web site that relied on 
scraping of another web site in order to work.  This was all legal, 
since the other company was the one actually paying us to scrape their 
site.  The contract specified an API, but after the deal was done they 
remembered that they didn't have one.

Anyway, one day our scraper site broke.  I was told that they changed 
"every page on their site".  When giving my estimate, I was clear that I 
had not so much as looked at the scraped site yet to verify exactly what 
changes would need to be made, but I gave an estimate of 1 week to fix it.

My boss went to the CTO and told him it would take 2 weeks.  The CTO 
talked to his liaison at the other company and told him it would take us 
a month to fix it.  They freaked out, and said there's no way it should 
take that long, and next thing I know my boss is telling me to give him 
a better estimate.

It turns out that all they did was change the URLs to have /en/ at the 
front of them.  I fixed our scraper site in about an hour.

For some reason, the CTO wasn't happy with me.

I learned never to give an estimate based on what someone else tells you.


More information about the PLUG mailing list