How do you recruit enough 'top talent' for your company?

I would look for:

• Local user groups and meetups. (meetup.com)
• Niche blogs and websites: they usually attract people who love doing their work. (http://jobs.jsninja.com/, http://www.crunchboard.com/, http://careers.stackoverflow.com/, etc.)

What works for me is to build a network of potential hires before I need them. I have found that this is the single best way to attract quality talent. This serves two other valuable purposes:

1. You get a sense of what other smart people are working on.
2. Sometimes if you are not hiring, someone you know is and you can refer them to your network.

Referrals are also good but it seems that is not working out for you.

One other thing to think about is what part of the business you grow and what part you might outsource. I know that can sometimes be challenging to sort out but it is an effective method to get you through a tough hiring crunch or at least allow you to focus on the right type of hiring.

Good answers already; I especially like Jarie's suggestion to consider the in-house / out-house split carefully.

Well, the question is one of those evergreen issues; I don't think there is an easy way to get great hires. Here are my ideas for this (some of which I have not had an opportunity to test in real life yet) :

• Have a a great "Working Here" section on your website, where "great" means accurate, plainspoken, detailed and honest; strictly no BS. Write extensively about how you work, with which technologies you work, what methodologies (Scrum, XP, etc) you follow (if any). Be so honest to have markers and anti-markers for cultural fit.

• Have a great working environment, and talk about it. Do you have 20%-time; does your company sponsor any worthy causes, do you contribute to open source? Can developers at your company order any book or PC accessory that they need without manager approval? Talk about this.

• Do phone screening, it really has a good ROI. Even small companies can manage 10+ phone screens per position; and it really helps cut out bad applicants. Consider including "wild cards" into the phone screen even if they more or less failed the resume step, i.e. if someone "sticks" to your mind as interesting, then phone screen him even if he didn't really match the objective criteria. Maybe he's the gem in the rough.

• Call or email all thought leaders in your technology / business domain, and let them know you're hiring. Maybe a well-known blogger on the other side of the planet knows someone who lives right next door to you.

• Consider using (good) headhunters. I have worked with 2 good headhunters, and they are really worth their (outrageous) salaries. Headhunters have a very wide network, and they do get attention when they call -- it is flattering when a well known headhunting company calls you. To some extent headhunters can get away with things I cant. For example if I invited a key employee of a direct competitor out for lunch it would be seen as very confrontational; but a headhunter could get away with it.

• Consider building an internal mentoring pipeline. From business literature it seems that the only long-term solution is to get get people with great potential while they're young (fresh out of university or close), and then form them to your needs. I believe talent for programming is something you can screen for; and most patterns and project methodologies can be taught within a few weeks.

I don't think there is a 'silver bullet', something that gives a 10x improvement for the hiring process. Or at least I don't know of one. It's mostly the same old venues everyone uses; past colleagues, networking, job boards (general & specialized), adverts.

I LOVE this subject. I am an operation guy, so human capital issues are on top of my mind every day.

Here is what I have learned in 15+ years in startups:

1. It all starts with culture. Go past the culture of tchotchke "benefits" like in office Guitar Hero setups, foosball tables, and other stuff that makes your culture fit for only one very narrow demographic group. (see my last article called "How a foosball table can kill your startup – part two"

2. Build diverse team from day one, not when you are 20+ person company. We have done it in one of the software development companies I was in. I joined them as #20 and by the time I left (2 years later) we were 120 strong with zero retention problems. We attracted and developed some cream of the crop people and white, single, 20-somethings were not even 25% of our company. I have two articles on the subject: "How to build your startup core team" and "5 ways to build diversity into teams"

3. Fish where others don't. E.g. you can find some amazing UI/UX people who are graduating as psychology majors (too many "fish" in graphic departments), great developers could be found in music schools (I've met some great devs who came out of Juilliard and Berklee)

4. Stop depending on "rock stars". They usually are exactly like rock stars - needy, overpaid, moody, unreliable, and will hop to the next place for couple of bucks more. Much better strategy is to look for people with absolute passion for their craft, insaciable hunger to learn, and humility. You can mold a humble smart person into a true treasure. You can't mold and asshole into anything but an asshole.

At the end of the day it will take your people raving about your company to attract really smart people. Smart people love working with other smart people.

UPDATE: 7/3/12. I keep on seeing folks coming to my blog from this answer and just wanted to add one more post you may find helpful: How To Hire Top Talent Without Paying Recruiters

It is unclear whether the nature of your business requires on-site presence. If it does not, go global. I know a successful software company that has no office altogether - all employees and contractors are working from their homes using virtual desktop technology. Need I mention that they are geographically distributed all over the world?

• Start an internship program. If there is a lot of competition in your area, do something different like a summer program for 2nd year students instead of waiting for graduates.
• Join or start a developer's group in your area (meetup.com).
• Market your company towards developers as well as your customer base.

I know you're not trying to make a pitch for your company on this site, but you haven't given any reason(s) why a top developer would want to work for you.

I would think twice before go in a quest to fill my company with Rock Stars. Not only its going to be quite a pain to manage them later (there's a strange correlation between skill and social disorder), as its probably going to be a waste of resources.

Unless you're building a quantum virtual machine, you'll hardly need a team full of rock stars. A plain, hard-work fella with some experience and skill is good enough most of the time. So, don't spend time or money to search and contract a Paul McCartney if a Norah Jones is just what you need.

Generally, a team good enough for you needs has 60-70% of experienced dudes, 40-30% of beginners (for more simple, not-creative tasks) and only 10-0% of rock stars.

Allow (and give then incentive to do so) your most charismatic collaborators to give lectures in some very well targeted universities/diplomas.

I am managing a dual education engineering degree and a MsC in complex systems architecture, and trust me, the most clever students have firm job offers way before leaving the university, thanks to some of our (external) lecturers. As a teacher, you can see the best potentials, and you can mentor them in order to make sure they will acquire the skills needed for being a good collaborator in your team.

I endorse Brunno Silva's answer. Great stuff. Also -

You need to define what you mean by "qualified" or "top talent". And you need to tell us your expectations. Is this an overall capability thing? Or are you looking for perfect experience? Are you looking for very particular niche experience? Or have you defined your preferred candidate in terms of being exactly like the competitions' people? Or do you expect to optimize all of these factors?

Now for something that will make half the people reading LOL or sneer - hopefully not:

Perhaps you should court and be open to the invisible but large segment of unemployed or poorly employed IT workers. Especially the segment that is scorned by all hiring parties - the highly experienced, veteran (15 years+) IT worker.

The startup and social networking segments always appear to ignore the vast swaths of laid off, capable, senior IT people located outside major technology centers of the US. MIS, telecomm, and in some instances insurance and financial industry types are all hurting for work.

What exactly is the matter with finding a laid off client-server developer (yes, older stuff from 8-15 years ago) and working out a retraining deal?

"Past" technology experience commonly appears to be interpreted as a taint of badness by those immersed in new technology. That is an incorrect judgement that reflects ignorance. Generally the current technology stacks used in the web and social media industry and by most startups are far simpler than the garbage that much of the senior technology crowd has dealt with in the past - such as Frankenstinian client/server and early web technologies.

Truly excellent past companies such as HP, DEC, and IBM in their primes hired for capability and character, and then trained for specifics. Today, the transition for someone to go from (say) Perl to Python or RoR appears trifling. (Anyone honest about it knows that this is the case.) Whereas a developer familiar only with current tech would strangle on older technology's quirks.

I did not even mention that someone who has seen a few fads come and go in their career will tend to be more flexible when the tech changes again.

Today companies expect perfect fit right out the chute. No wonder finding the "right" person by that criteria is virtually impossible.

The right person may have the "wrong" current resume, is what I am saying.

A company I worked with had an interesting solution. As well as developing software they ran a recruiting business on the side, placing people into contract jobs. If a contractor proved to be very good they hired them in-house.