HNHacker News
TopNewBestAskShowJobs

seanjohnson

295 karma · joined November 24, 2010

submissionscomments
seanjohnson··on Funnelcakes: tasty ideas for growing your business
Great suggestion - added an example to the homepage.
seanjohnson··on Always find the best candidates - the 7 minute candidate matching system
Interview street looks fantastic. You probably don't need a system like this for dev hires with a tool like that.

And you're right about designers - I look at portfolios more than anything else. This system does save me having to look at resumes to figure out whether they have a portfolio or not. It at least weeds out that third :)

seanjohnson··on Always find the best candidates - the 7 minute candidate matching system
What a screening system like this does in practice is save you from having to look at hundreds of resumes from people who aren't even remotely qualified - the folks who carpet bomb their resume out to anyone.

You're the one creating the criteria and determining the weighting, so if the system excludes people because they don't have J2SE experience it's only because you put it on the list and made it extremely important.

seanjohnson··on Always find the best candidates - the 7 minute candidate matching system
It is binary, mainly because it has to be. You couldn't reliably job out a screening like this if it required the person to understand the nuances of JS, or any other skill requiring domain expertise.

But as long as you weight it properly, any candidate that makes the short list has javascript as competency. Their level of expertise can be teased out as you go. This is meant to be a first step to go from 300 down to 20, not to take you all the way there.

There are other ways to make it more nuanced - identifying keywords that would make it more likely to identify a JS expert in a resume or cover letter (not a replacement for looking at someone's code, but again something a non programmer could do). Or looking at someone's github account and seeing if they have JS repos (maybe factoring in the # of people following those repos). The formula would need some tweaking, but those are objective measurements that could get you closer.

seanjohnson··on Always find the best candidates - the 7 minute candidate matching system
The point of the system is to enable someone who doesn't have your knowledge to do the initial screening with a high degree of accuracy. The weighting of the criteria is usually what prevents someone from getting tossed for the wrong reasons.

- The college degree example is solved by weighting, as I mentioned. I agree with you - we've hired plenty of people in the past who didn't have a degree. That's why it would be considered secondary criteria - if you do it right they won't get heavily penalized for not having a degree if that's not important to you.

- LinkedIn, portfolios, etc. - the point is again what's important to you. For example, not giving me a portfolio is a deal breaker. I have hundreds of resumes to look through, and if I can't easily see your work I don't have time to hunt for it. It says at a minimum you didn't take the time follow instructions in the posting.

- The third example is your strongest argument. It can also be solved by weighting though, by giving more difficult technical skills more weighting. I know plenty of people who look for folks that know Java because it's relatively easy for those folks to pick up Ruby.

You're right - a few diamonds might slip through the cracks. But in most cases I think you'd miss them anyway.

seanjohnson··on Always find the best candidates - the 7 minute candidate matching system
impossible :)
seanjohnson··on Win customers and improve products with HTML prototypes
That's very clever. Thanks for suggesting - we'll have to try that soon!
seanjohnson··on Win customers and improve products with HTML prototypes
That's pretty slick man. Nice work!
seanjohnson··on Win customers and improve products with HTML prototypes
Agree - we primarily do product development, so it's usually relevant for us. I will say we've implemented HTML prototypes for checkout processes in the past with some success as well.

Can you elaborate on how you've used them to set scope?

seanjohnson··on Win customers and improve products with HTML prototypes
Never have - looks interesting though. We'll have to try it out!
seanjohnson··on Win customers and improve products with HTML prototypes
Haven't been able to convince ourselves to use Comic Sans though :)
seanjohnson··on Win customers and improve products with HTML prototypes
Totally agree - other than relevant images (the shoes, in this example) we try to avoid using much in the way of color, icons or other design. We use variations of gray shading, unstyled buttons, Helvetica across the board.