CosmoGTD said:This is exactly what I just went through with my recent project.
I really thought it would take 5 hours, and it took 16 hours! That blows my mind I could be off by 3x.
This is a very common occurrence. Some part of our brain knows its a 16 hour job, whereas we think consciously its a 5 hour job. No wonder we resist it.
I think this is why some people recommend breaking projects down into actions that will take an hour or less (some say 30 minutes or less) to finish. The more granularity you have, the better the chance that you've estimated the time accurately.
My own solution to the problem has been to keep track of how long things take. I don't bill by the hour, but keep track of time as if I did. That helps me see how much time gets chewed up by administrative tasks or marginally useful stuff (like the Web!), and how much is actually available for client work. At the end of a project, I can compare how long I thought it would take against how long it actually took and (hopefully!) improve my estimates for the next project.