Ask HN: Why do the collaboration websites always fail? | Hacker News Reader

Ask HN: Why do the collaboration websites always fail?

9 pointsSkywing9 comments

I know the topic is a little vague, but I didn't want to make it a super long one.

Everyone has seen those websites that try to connect developers with other developers based on projects they want to work on. There have been many attempts at them. I don't think any have been terribly successful, though. I could be wrong, of course. I'm not really talking about elance.com, or monster.com - those are more freelancers strictly looking to get paid for their services. I'm talking more like openinvo.com, except without the upfront fees.

Why do most of these sites seem to have such a tough time being successful? It's definitely something I feel most developers or web designers, etc, would find useful. I know I always hear people complain about how tough it is to find people to work with on ideas of theirs.

I think most of these types of sites end up becoming just a deadpool of "want ads". It's not very fulfilling to post an ad saying you're looking for a project to join, or post an ad saying you're looking for people to join your current project.

So what do you think? How can these types of services be made better?

One hypothesis (not sure if it's correct) is that there isn't actually as much of a matching problem as people think, but instead more of a supply problem. When you're a lonely open-source developer who can't find a second or third contributor, if this hypothesis is right, then it's not because they're out there, itching to write code, but unable to find your project in the noise of the web. Rather, it's because they don't exist, or their time is already spoken for. If that's true, a matching marketplace will just result in a bunch of unanswered want ads from lonely open-source developers, because the supply of willing hands is already taken. At best, it'll result in a zero-sum switching of developers from some projects to others.

It seems for that scenario not to be the case, either: 1) there has to be a bunch of unused latent supply due to a real matching problem, i.e. lots of people who would contribute but cannot find anything to contribute to; or 2) the existence of such a site would have to actually increase supply itself, perhaps if it generated a bunch of positive PR and reached previously unengaged people.

I think github is a "de facto" collaboration site, but the process is different... You start working on something you find interesting, and if others feel the same way, they might contribute.

Great point. Github pretty much is the collaboration process, just taken to an entirely online format.

I agree, but Github is a tool, it doesn't have an easy way to discover projects to work on. This is why I started this site http://hwknd.com that you can use to post your new github projects and find collaborators. It's still in early steps, but its own code is on github too so you can build on it if you like.

I agree completely with this. The best way to find other people to collaborate with on a personal programming project is to do something other people want to do. Talk is cheap. Take a look at the number of "projects" on sourceforge that consist of a name and a description and haven't released a single byte of code. I'd be willing to bet they outnumber real projects.

Haha, I own one of those. Code's been sitting on my drive for 5 years, but it's far from ready enough to be uploaded. I'm surprised SourceForge doesn't purge the empty projects once in a while to free up project names.

Obviously you've got the problem of people that are good at doing things being fewer in number and less likely to need to actively hunt for collaborators than people looking for others to do things for them, which skews the typical site visitor towards those that post ads rather than reply to them.

This could be a problem for StackOverflow etc as well, but the difference is that people are both happier giving free advice to strangers than embarking on large scale projects with them, and more likely to stick around and answer a few questions on a site that might help them on frequent occasions in future with other issues.

Even if they work, sites catering for people making big one-time decisions like finding a cofounder are always going to end up looking like a ghetto. The people that find what they're looking for don't need to come back.

Not all of them fail, its just that people like their collaboration to be 'thin', not 'thick'. Stackoverflow is a nice example of a success story. Casual, 'thin', very effective.

People are pretty set in their ways and by forcing them to completely shift to some new platform the majority of the collaboration platforms misses the point that we don't actually want to do that. Collaboration is more about communication than it is about anything else so whenever a collaboration site is successful it usually is because it leaves people free and solves a communications issue.

I think the reason is that internet is full of trolls and if you don't put significant effort into making community / collaboration clean and vibrant then the collaboration site is destined to fail.

For example, compare Quora and Yahoo! Answers. I don't thing Quora is successful because they have very good UI (it does not hurt), I think they are successful becasue they are very serious about making they site clean and without trolls. On other hand, Yahoo! Answers is 'troll heaven'.

Reply on news.ycombinator.com