Yes, hiring people who are consultants/contractors have the advantage that you can “try before you buy” with the drawback that I've found that people with those backgrounds are often very wide but shallow. More _ shaped than T

We've found one thing that works is having people fly in on a Friday and do a 3-day interview crossing into the weekend, and after 3 days, you have an amazing sense for that person's skillsets.

One of the greatest things about oDesk et al is the ease in which you can do the 'working interviews' and scale from 1 day to 5 day to 3 months to valued full time employee. Sure there are a lot of downsides as well, but this fact nets all of them off and then some.

Also, I know you've been writing this blog (or working in viral marketing/metrics) too long when you write “the depth gives them a source of knowledge that makes them viral to the team.”

Shouldn't that be 'vital' instead of 'viral'. I am not sure you want a team member going all 'viral' on everybody else

Unfortunately this trait is often well hidden during a typical interview, and only appears later.

I am involved in a social network startup with the freemium model and I wanted to ask you where can i find social network historical data or statistics like virality or free to premium conversion, to backup my business plan data!

Thanks for all your essays!

Ignacio

1. Hire people who already do the job as a hobby. This is a silver bullet. At Firepad, we mostly hired coders who had shipped their own PalmOS shareware. They were easy to find, knew the API, but more critically, they had proven motivation completely finish Palm software apps, without any oversight or schedule pressure.

2. Read at least 10 resumes for every interview; interview at least 10 for every hire. It's a numbers game. You need volume, but have limited time. Face time is costly. Narrow the field.

3. Use an intern to take a first cut at resumes. For each job, define features to look for, and have them present you with the winners. Then you have a stronger pool on which to execute #2 above.

4. Interviews: rate candidates on a scale of 1 to 3 for motivation, integrity and ability. No fractions — if they aren't a 3, then they are a 2. Any candidate scoring a 1 is rejected. Any candidate with more than one 2 is rejected. The very structure of this system forces hard judgment calls to be made quickly.

5. Hire someone just for a day. If they perform, they get the job. If not, you pay for their time and show them the door. Interviews are essentially hypothetical, while a one-day contractor shows you plenty. This requires a slack labor market (e.g. 2002).

6. Either fire your new hire within 2 weeks, or never. Why only 2 weeks? First because, if they aren't great then, while on best behavior, then they never will be. Second because, if you fire later, after friendships have formed, the whole company will suffer bad feelings. Hesitating to reverse a mistake is probably the most common, yet most destructive, hiring error.

Credit where due: learned #4 and #5 from other sharp software company founders.