In December last year (2012) an inner voice told me to be more active on Stack Overflow. More active? Ok, active at all. Up to then I had almost exclusively asked questions rather than answered other people’s questions.

I quickly got the hang of it and challenged myself every day to do more. Until I got tired of it. Here’s what I learned in the process.

Be successful on Stack Overflow?

What does it mean to be successful on Stack Overflow? At first it seems easy…You are successful

• if you ask a question and at least one of the answers solves your problem
• if you answer an question and it is accepted

Or are you only considered successful if your reputation is higher than…100k? 20k? 10k? I guess it highly depends on your motivation for hanging out on SO in the first place.

Initially, I was interested in finding out how quickly I could push my reputation from then ~500 to 1’000. It was kind of an experiment. Once I got there I reflected a bit and questioned my own motives. Clearly I wasn’t doing it to improve my karma for being so altruistic. Seeing your reputation rise feels good, it gives you a sense of achievement.

BUT answering questions is rewarding in a different sense, too. You learn a lot! I happened numerous times that by analyzing the problem (that lead to the question) I picked up valuable pieces of information from docs I was going to quote or link to.

How to get reputation fast

As a representative for many similar questions let me quote from http://meta.stackoverflow.com/a/17250 (follow link for details) as for how to get reputation fast:

a) Post quality answers.

b) Monitor the frontpage and the new questions list.

c) Setup a good but short list of Interesting and Ignored tags.

d) Avoid Wall of Text questions.

e) Post an answer even though the question has 1–2 or even 3 answers.

f) Learn when to edit. Post a short answer at first and then edit.

g) Be humble, thorough and fair.

If you want to be successful on Stack Overflow (again, whatever that means for you), I believe that’s a valuable list of proper recommendations for how to do it right (list items reference those in the quote).

a) Posting quality answers obviously takes time for most questions. However, if you spend time studying the question and post an answer anyway you might as well spend a little more to do it properly. Sometimes it’s easy, though. Many times all it takes is a quote from an official documentation and a link to it. See “Is Stack Overflow the Google for programmers?” below.

b) Monitoring questions can be complemented with email subscriptions for tags you’re interested in. For a while I subscribed to ‘java’, ‘spring’, and ‘spring security’ but when I got the respective email, every 3h 6h or 24h, I subconsciously felt enormous pressure to go through them all. This was so time consuming that I stopped after a week or so. Many times I’d get the n + 1st email before I even had the chance to look at email n.

e) I disagree, I don’t wast my time on such questions.

f) This really helps – if others do it, too. If I see that a new answer was posted while I’m writing an answer myself I open it immediately. If the answer, or the answer skeleton, is written by one of the well-know Stack Overflow premium league authors and it appears to be plausible (which it usually is) I stop right there.

However, if you’re primarily after building reputation i.e. if that’s your measure for success one thing is missing. You need to assess if a question is worthing “investing in”.

How to pick the right questions to answer

• I hardly ever answer down-voted questions
• I usually only glance at questions by users with a ‘userNNNN’ name. If you don’t bother to pick a proper username, why should I bother answering your questions?
• I take the reputation of the user into consideration. WARNING, what Stack Overflow displays here is not the user asking a question but the user who last submitted an answer or edited the question!
• Questions with answers, particularly those with accepted answers (the ones with the yellow font on the olivish background square) need to sound tremendously interesting to catch my attention.
• Ideal candidates are questions with 0 answers but many views. Even better would be if such a question had a few up votes, too.

Is Stack Overflow the Google for programmers?

An odd question I know…obviously programmers come to Stack Overflow looking for answers, after all that is its raison d’être. However, I feel that an increasing number users seem to neither check Google nor Stack Overflow first before posting a new question. Compared to Google the Stack Overflow internal search is really poor. I usually start with Google, the relevant SO question normally rank pretty high.

Stack Overflow is brilliant in suggesting existing questions once you start formulating your question. Many new users seem to even ignore that. That really annoys me. So, instead of searching (Google) first n00bs type in badly crafted questions.

Dudes, SO is complementary to Google and not a its replacements!

Personal statistics

The below reputation graph clearly shows a peak around beginning of January ’13.

My SO flair reporting the current reputation and all the badges I “earned”.