This and the other startups like Interview Street seem to be focusing on the less lucrative side of the hiring problem. Like Joel Spolsky said, and like I know from personal experience, my best developer friends aren't in the market salivating to take interview tests, and do code sprints to prove they are good. They are passionate about developing, have many great jobs to pick from at a given time if they tried. If they have free time, they would rather put it on an open source project that excites them, or a smartphone/web app that they are making for themselves.<p>If you look at filtering mechanisms themselves, there are companies that try to have puzzles that they ask you to solve before joining (eg: ITA software). But they are really challenging puzzles that I would solve just for fun anyway, (which is what I did for months till I one day decided to apply for a job there). The fact that these puzzles are also made by the company (and not a third party website) makes a difference. It tells me about the people that work there and potentially about their culture. On the other hand, building a random web app to prove basic skills is not something I would imagine most talented web developers would like to do anyway.<p>If you look at it from the perspective of a company that is hiring, tools like this solve merely one part of the puzzle. It takes the subset of developers who are already in the market for a new job and filters them. As a startup though, what I want to do is lure the great engineers who are content at their regular jobs and not looking at all because of inertia. It sounds like there is more of a "search for developers" problem that will expand the breadth of all the developers available to me, instead of focusing on the ones who are already out there (and are most likely not the cream of the crop by definition).<p>In short, I don't mean to be negative, and am not saying that there might not be a market for a tool like this. But I am pretty certain that if you are a great developer, you don't have an incentive to make a web app on this.
OK as someone looking for employment, I feel ambivalent about this and probably would pull the plug on the process if you asked for non trivial work from me. I understand the need to weed out applicants that lack skills. This can and should be done by technical screening conducted by qualified individuals.<p>FWIW this opinion is based on experience. I've conducted more interviews than I have given. As an applicant I once went through a fairly lengthy interview process that first did a screening interviews including live coding exercises. They were conducted and critiqued in real time. I have no issue with these. You can either do what you say, or you can't and it becomes apparent fast.<p>The next stage is where I started to question the company's processes - They requested a small "project" followed by an interview. I _nearly_ pulled the plug on this, since it was a request for 10+ hours coding, but did it anyway.<p>The next interview involved a fairly in depth defense of my code and design choices, white board work that again put my skill set to the test real time. After several hours of this, I was requested to do another "small project" and I walked away.<p>TL;DR Ask me a million questions, put my skills to the test on the spot, but don't ask free work... I'll think you're unable to judge my skill/confused/a scammer or whatever.
I first posted a snarky comment saying how this is spec work and what not. I then looked closely and deleted it, because it's not a spec work.<p>Your description -<p><pre><code> Recruit web developers by having them build a web app
</code></pre>
should really be this instead -<p><pre><code> Recruit web developers by having them build a TEST web app
</code></pre>
You have to be <i>very</i> careful with how you describe your service. You are an inch away from a much despised domain area of 99designs, and it's only sensible to try and distantiate yourselves from any spec work associations.
Hi HNers,
OP here. Cull.io is similar to top coder or interviewstreet but focussing just on recruiting web developers. It works by asking candidates to develop a web app on localhost instead of presenting a textarea where candidates types the code.<p>Advantages:<p>1) Candidates can virtually choose any language, framework, libraries or IDEs to solve the challenge. (Since they are developing on localhost with their own setup)<p>2) Because of 1) you need not limit your challenge to be overly algorithmic. You can for example ask questions like 'Pick the most popular article on hacker news today'<p>Disadvantages:
Limited to candidates who can develop a web application
Leaving aside whether or not this is a valid way to attract top people, my suggestion is to rebrand the name - "cull" has too negative a connotation ("cull the herd").<p>So #1, I'm part of "the herd", and #2 this tool is being used to possibly cull me. Great. No thanks, I'll spend my time applying to a place that doesn't have a tool dedicated to weeding me out. It doesn't portray the hiring org or your tool very well.<p>I'd be afraid that you're turning off devs right away, before they even get to the site (which has the potential to turn them off even more). My $0.02
I think you'll have traction problems on the developer side.<p>Doesn't matter what kind of job it is, you can be pretty much 100% guaranteed that the job will be listed on the company website without any of the extra hoops to jump through.<p>Requiring someone to build a web app just to apply is a huge requirement. It's fine if you gave someone an interview, and this is the final step before you hire...but requiring someone to code just to submit an application is abuse.
A lot of people are commenting that they don't want to have to solve this type of problem before applying, but I don't see an issue. I could whip up a simple program that listens for HTTP requests and e.g. reverses a string in minutes, using any high-level language. I spend more time than that writing a nice cover letter.<p>Given the number of "non-coding coders," I think this is a decent service.
Seeing things like this is why I love working in open source, a quick look at github + the projects I am involved with and someone can see what my interests are, what my abilities are and how I work in a team.<p>Not every work place produces open source code and not everyone wants to spend their limited spare time 'working', but I cant think of better advice for eager job seekers / students / graduates to look for open source projects to contribute to.<p>(of course it isnt the only reason, making software you love, with awesome people, and making the world a better place arent to be ignored)
I think it is important for some cases at least, that the libraries used by a candidate be limited. For some algorithms (lets say a search algorithm), it would be convenient to use a searching library. What library a person chooses is informative, but not as much as how a person implements it himself/herself. I do understand that this is essentially a screening test, but if lets say, DuckDuckGo was hiring, then how a person implements searching, even if trivial goes a long way in deciding how applicable that person would be for the job. This will be extremely difficult to implement though, I'm just saying one can require them to submit source code and then manual checking can be done for candidates who pass the screen. The reason is that there are a lot of drag-and-drop style wizards that can just generate some skeletal web service that could be used to break this screening test.<p>In the end, this is an amazing service, but sometimes, these tasks are essentially trivial, like, for example, exposing a function to a web service. I believe (and this is just my opinion), that the function is more essentially than how one can wire it to different interfaces.<p>But, what do I know, I'm too drunk right now.
There's definitely some merit in this approach over something like Fizzbuzz and whiteboard problems. At my previous place of employment [1] , after the initial interview, you were given 48 hours to produce a simple web addressbook app. You could use <i>anything</i> that you wanted to produce the app - at the end of the 48 hour period, you had to supply thme with the code and instructions on how to install it on a Ubuntu instance. It was actually quite fun (and then after getting the job and testing other people's entries, you begin to see how it separated people with impressive CVs who couldn't program for toffee, and others who had much less experience, but managed to come up with something that worked!)<p>[1] That would be Open Source Integrators, a great Durham-based outfit who I believe are still looking for people to staff their Chicago offices: <a href="http://osintegrators.com/" rel="nofollow">http://osintegrators.com/</a>
@ksat - I think it's easy to get the wrong idea of what your site is for from the heading - "...having them develop a real web application". Of course building an entire web app just to apply for a job is overkill. But it looks like that's not what you meant? I read further down the page and the examples given are more like short coding challenges. That's acceptable, and I think you need to focus your site more on filling that need. Position it as some sort of "Automated Fizzbuzz Test".
What stops me from asking a friend to help me? Or worse still, if this is used as a campus recruitment tool, someone might mail working code to the entire college!
I have not seen the service itself for the developer, but as for the screen I would recommend a way of customizing a bit more what site it should check. Right now it seems you can only change the port, but plenty of times I just create a new directory under root for the webapp (something like <a href="http://localhost:8080/mywebapp" rel="nofollow">http://localhost:8080/mywebapp</a>) because it's easier in some web servers.
Oh, now programmers are graphic designers too? Don't want to play the submit three logo comps and maybe we'll hire you game?<p>Geez, even a hired gun has to shoot someone for free once in a while just to show potential employers he's still got it! No one really knows what you did before (unless you have code groupies). Employers certainly do not give a shit in today's job market, they want to know you will fit in with the team and that you work as advertised. No one in their right mind believes advertising or resumes these days. So what's the alternative? You want to slip them source someone else paid you to make! I sure as shit would not hire you if you share like that. Competitions surely suck for losers who do not learn from their loss(es).<p>Suck on this - if you are bored then YOU are boring. If you won't build a small (not going to make anyone else money) solution to show a potential employer you can, you probably can't.
it better be a very simple web app.
otherwise the applicant can only apply to a couple jobs a week/month.<p>makes me think: ask a car manufacturer to design & build a new car before you choose them to design & build your company's custom car.