First of all, I built the software only after I found 3 people in my circle of friends who were willing to pay it forward. This allowed me to understand if I was solving a real problem people would pay for or not (it was important to me, but this might not be of the same importance if you are on a freemium model, which might require a different take on this).
After that, I pretty much used:
1. Referral (for every client, try to make them super-happy every time and then ask for a referral of them)
2. Networking. Onion2k already told this, so, join and meet people who might help you or need you. In my case one breakthrough was to find out my product featured on product hunt (and near 100 upvotes) thanks to the relationships I built over time.
3. Reddit could be another place you find useful, although in my case I discovered that it's a community more geared towards free product (so if you have a trial, you must solve a really powerful problem to have them try it, or have a super-compelling homepage).
4. Quora is another place I gained a bit of traction thanks to some small, targeted questions I answered. Not much a big deal, but it's worth checking out.
Always remember, in everything you do, your goal must be to add value to the people or the conversation. I personally think that when you do only for marketing purpouse and not for the sake of adding more value to what you see, everything ends up being bad.
5. Giveaways are another great idea you can use, find a good blog/magazine in your niche and see if you get to place a giveaway over there.
6. If you are still struggling on how to build your network, the easiest way is to help people. Find people in your niche and help them in any possible and honest way. Don't think about marketing, think about just helping them without selling. Networking is made of friends (at least to me).
7. Timing is everything: If your website is featured in something like Product Hunt, JOIN the conversation and if you can, offer a greeting offer on your product page.
Also... time and effort is required. LOTS of it. For example Airbnb didn't grow up in a day, but grew up thanks to small, continuous ideas and efforts. As toumhi told you: do not automate until it's needed.
If you are an engineer you will want to automate. Resist that temptation, do it only when you can't do it yourself at all.
One nice read about gettin traction is tractionbook.com . I read it recently, and while some of the channels they describe for getting early users might have a price, others are really free, so it's a worth reading.