How do you overcome 'Analysis Paralysis'

Question

Time and again I find when researching something completely unknown, I go down so many bypaths that lead nowhere. For example, I was searching for a UI Automation Test tool... Stumbled onto Project White... moved to UI Automation Toolkit... to a Bugslayer article wrapper for the same... thought it would be a good idea to write one of my own... In the end nothing got done except may be I'm enriched by the knowledge. :)

It all started with automating a Smoke Test... which could be done easily using QTP which my company has licenses of (but no one at the moement to write the cases. BTW, I do not have access to it).

I'm sure this is a road well traversed by my fellow journeymen.

Advise?

Answer 1

Make a little project to test out how well each of a few different ideas works, as some software will have a trial period so you could test it out. See how well various software performs and at the end consider writing a recommendation, either why they all do not meet your minimum requirements or which of those that do appears to be the best value for the business to buy and change processes to use.

I went through something like this when testing out various AJAX frameworks a few years ago for an ASP.Net web application just before Microsoft came out with their own implementation and spent a couple of weeks trying out a few different ideas and chose one in the end and did the integration which was rather simple as I had already spent some hours toying with it.

Answer 2

1. Be clear about your goal
2. Give yourself a deadline
3. Like somebody else said, write small programs to test ideas.
4. Make decisions quickly.

Don't worry about bad decisions,just trust yourself to figure them a solution when you get there.

Answer 3

Start off by reading as much as you can on the subject by sources you respect, acknowledging that there is a lot of BS out there, but a commonality of positive experiences is a good sign. Then narrow it way down to just a few possibilities and then write small suites of test code to prototype the concept, test the API, etc. Then if nothing pre-packaged will work, consider writing it yourself. The main thing is to try to narrow down the field of possibilities with the initial research so you don't get bogged down trying too many things.

Answer 4

Let's be also aware of the fact that "analysis paralysis" quite often means the same as procrastination. An sophisticated way of procrastination sort of. It's a handy excuse to keep us from starting with the actual work. At least that what I have felt a couple of times. If I had used half of my research time for trying to code something, probably half of the application would have been written already.

That said I have to add that social scientists recently discovered a strong link between procrastination and perfectionism. So it's not the lazy ppl who postpone, it's those who want to end up with the perfect result.

This means one way of overcoming analysis paralysis shoud be lowering our expectations towards ourselves.