Ask HN: How do you start a website that requires user contribution?

[stepanbujnak]

How do you attract users to use your website, if the website requires users to create the content? When there is no content, there will be no users. When there are no users, there will be no content. What are the best practices to escape this loop and build up a user base that is most likely to visit repeatedly?

Do you know/Do you remember what StackExchange looked like in its early days? How about Quora, Wikipedia or any other website with similar experience?

[wuliwong]

One story I've heard is from Alexis Ohanian one of the founderes of reddit.com. He said that in the beginning they created multiple accounts and posted lots of content themselves. So basically, they faked it for a while. :)

I believe the story with Yelp is that they paid people to write reviews in the beginning.

I definitely feel you, I have had similar struggles. I made a site that had some limited traction, mostly my friends used it. I think if your close friends are actually in the demographic you are looking for, then that is a great place to start.

Something else I try to consider is to make a site that strives to serve millions of users but can be useful to far less. I've been working on a locally focused, social network and I am creating it in a way that even if 10 people are using it, as long as they goto the same places, it should be useful. 10 people in the same area would be worth 1000 people strewn across the globe in my case.

But I do believe that it is a problem that has a solution which differs greatly from case to case. It's certainly a chance to show your creativity and hustle, I suppose. :)

[epsylon]

Steve Huffman talks about this in his Udacity course on web development: https://www.youtube.com/watch?v=zmeDzx4SUME

(The whole unit 7 talks a lot about how they built Reddit.)

[simantel]

Unless you have an existing audience you've built up before launching, I think the not-so-secret secret of most user generated content sites is that they faked it when they were starting out.

Alexis Ohanian has talked about how they seeded Reddit with posts made under various fake usernames for the first few months after launching. They made it so that as admins, they just had an extra field for "username to post with" when creating new posts.

[Tmmrn]

https://www.youtube.com/watch?v=zmeDzx4SUME

[jackgavigan]

One way is to pay users to generate content.

Back in 1999/2000, I was CTO of a startup building a website aimed at independent travellers (which we defined as anyone who wasn't on a package holiday - i.e. people who were booking their own flights and hotels, and picking their own activities). Our plan was to become something like what Wikitravel is today - a kind of crowdsourced travel guide, with recommendations from other travellers for places to eat/drink/stay, things to do, sights to see, etc.

The idea was that a traveller would come to our site, search for, say, places to stay in Dublin, and the list they would get back would consist of recommendations/reviews submitted by other travellers.

We faced the exact same chicken-and-egg problem the OP describes - we had a website based on user-generated content, with no content.

People who were travelling/backpacking around the world during their gap year were a clear part of our target market and, at the time, there was no mobile internet, so gap year travellers would use internet cafés to access their email. The biggest and most popular internet cafés in London were run by easyEverything, where, on any given day, you could find dozens of travellers and tourists.

So, we did a deal with easyEverything to buy terminal usage in bulk, and we then gave away £1 worth* of usage to anyone who wanted it, in return for them writing a piece of content for the website.

We literally had people, wearing branded t-shirts, at these cafés, grabbing people as they came in and asking them if they wanted some free Internet time. From our perspective, it was a very successful approach - we were getting UGC for less than £1 per piece of content and, at the same time, publicising our site and getting a bit of brand awareness.

Unfortunately, the company didn't survive the Dot-com Crash but that's another story...

* I say "£1 worth" because easyEverything used demand pricing - if the café was empty, that £1 might buy you 90 minutes, if it was busy, it might only buy you 20 minutes.

[rwhitman]

You have to fake it. Yelp seeds every city they enter with paid copywriters for instance, Reddit was entirely fake accounts initially. Airbnb pays for photography of nicer listings.

Either hire professional copywriters or recruit unpaid off-site interns to do it for you. Personally the intern method I think works best because they are dedicated to giving you good content, but really rough around the edges and the content feels very genuine. The NYC craigslist jobs section works phenomenally for this.

[cdcarter]

Pay your interns. If you have any questions or concerns with a policy of paying the people who work for you, stop what you're doing now.

[mappum]

Will people really agree to an unpaid internship for some brand-new, unknown company? What's the reasoning there?

[rwhitman]

For those trying to get a foot in the door in competitive creative fields, an internship is an important first step in starting a career or pivoting to a new one.

[danbmil99]

> What's the reasoning there?

Startups! Kewl!!

[drwl]

for experience and an entry on a resume

[untilHellbanned]

How do you "fake it until you make it" with paid content? That is if you have a site where the user contributions generate revenue for the user. Totally unethically to create fake accounts and/or use somebody else's content to receive money for yourself.

[vii]

Having worked on a growth team for a few years and where I incremental millions more active users above trend, the underlying techniques are surprisingly simple.

- implement many different ways you can display the content you have and try to make it relevant

- don't display empty pages, try to show the most relevant content you have even if it's not great

- if you want people to perform an action (here, adding content), ask them to do it and make it easy :)

- send out email or other notifications, giving social context of other people performing the action you want them to perform and with a clear call to action, once again

[morisy]

The cynical way to say it is to "Fake it til you make it." The nicer way would be to say "Be the change you want to see in the world."

When we launched MuckRock, a user-generated FOIA request site, we had about 50 early registered users who were dying to use the site. Then we set them loose, and after 2 months, absolutely no one had filed. So I started filing a ton of requests on my own, got back really good information, and all the sudden users started filing.

When I went back and asked why they hadn't filed before, they said they simply didn't know what filing should look like, so having some examples was helpful.

We also found a few power users who were already doing this on their own, and gave them free premium accounts. We lost hundreds of dollars on them for the first year, since we covered cost of stamps, scanning, etc., but it more then made up for it in helping model the kind of community and content we wanted to have.

So the short answer is, find ways to bootstrap it yourselves, whether that's just you compulsively dog fooding your own site or paying users to ask questions (it's almost always harder to find good question askers than answerers, and the latter comes naturally to sites with good questions), or, like Wikipedia, intelligently and ethically scraping some bootstrapping material from existing resources.

[alexeichemenda]

You have to create the content by yourself. No user is going to create your content magically on an empty site. So build your content (it will be a rather long process), and then it will (not organically) generate interest.

Also, target your friends / people you know that could use your website, and ask them which content they would like to see, if they want to help you build the initial content.

What's your website ?

[aaronbrethorst]

This is what I had to do with http://www.cocoacontrols.com for about six months before I had enough UGC for it to be sustainable.

Finding acceptable repos on GitHub, downloading them, running the projects in the iOS simulator, screenshotting them, and writing up a description took a significant amount of time.

[wuliwong]

What is UGC?

[aaronbrethorst]

user generated content.

[wuliwong]

Ahhh cool. Thanks

[wuliwong]

Thats what I have been doing with my current website. It is a website for links for Swift. I have 300-500 daily visitors but haven't seen a large uptick in traction. Right now i easily have the largest collection of swift tutorials and references and my plan is to just stay the course and continue to build my repository. My hope is that there will come a tipping point in user traction during this growing of the repository of links. But thats just my theory. :)

[loumf]

A lot of people call this a chicken-egg problem. For some reason, that's supposed to make it seem hard to solve, when in real life, anybody could start a chicken farm very easily.

You simply buy chickens.

If your question is how do I start a chicken farm without buying chickens, then you have made the problem unnecessarily complex.

The thing is, though, when you buy a chicken, you can be relatively sure that you will get eggs. And you can be sure that if you have eggs, it's not too hard to get chicks.

Your real problem isn't how to get content (that's simple, you buy it), but how to then get users that want that content. And then, how to get some of those users to make content.

(to answer your direct question: StackExchange was seeded by JoelOnSoftware and CodingHorror readers -- it was a closed beta until they got enough content. In other words, Joel and Jeff already had chicken farms, and they donated a bunch of them to StackExchange)

[Mz]

This is called a "chicken and egg problem" and searching for that term will find you previous articles and discussions on HN. Here is a list of links I once put together: https://news.ycombinator.com/item?id=2126209

[tempestn]

Another approach is to make it backward compatible with something else. For example, if you wanted to make a new site/network similar to StackExchange (but presumably with some revolutionary changes; not just a "me too" copy), you could start by adding an interface layer using the StackExchange API and adding some of your new smarts to give your users some of the benefit of your tool, on top of the StackExchange content. Most likely you wouldn't be able to add all the new goodness there though, so you would also do as other comments have suggested to build up your own content. Maybe even incentivize users to add stuff based on StackExchange answers they find (assuming some manual intervention is necessary to make the answers suitable for your site - and also assuming you comply with terms of use and such).

It's hard to give exact examples without knowing what you plan to do of course, but that's the general idea. Make it backwards compatible with whatever the closest thing is out there already, so that even in cases where you don't have any of your own content to return for a user, you at least have something. Then do everything you can to ramp up your own content asap.

[balor123]

Does that work in most cases? Most communities don't provide APIs. A few will provide RSS feeds - not clear if you can legally reuse them like that?

[greenjello]

StackOverflow, Wikipedia, and other creative commons sites provide good starting data sets because all you need to do is attribute to them. No expensive licensing things, or anything like that.

Often you can write a pretty good bot using their data, and emulate user interactions. Enough users won't notice the site is run by bots that it's better than nothing if you don't have funding :-)

[kromodor]

It is important to keep in mind that the following is only an assumption. We'll be able to test it in few months when our website is live.

We have approached the situation from similar angle as many people here already described.

However, we involve a tool which allows people to make "cool albums" and can easy share it outside the site, into their existing communities.

We assume that since we will have almost none real users (beside the fake ones), inside action will be non-existent. Thus if the users use the tool to share outside, we might benefit from the outsiders checkout the album of their mates.

Of course, while this sounds clever, it hides risks - the early adopters should provide at least like-able content. And this, by experience, is not the general case. To be counted, also, is that our early adopter is someone who is already member of an existing community.

And whoever registers in the site should be carefully caressed to feel special in order to, at least, provide feedback.

tl;dr - fake accounts; a lot of personal attention from founders; people can benefit from the site without inside users.

[kidlogic]

Platforms require two important factors - self-marketing and publishing tools. It's important that the published content benefits everyone within the community, which helps build up to true Network Effects.

Github (same-sided platform): publish your repo, users can download and contribute to your code. More code, more branches, repeat.

Youtube (two-sided platform): Published content provides entertainment. Users can respond to others video to provide additional relevant content. Each publisher promotes their own channel for recognition (and in turn, promotes your platform).

I'd recommend reading: http://platformed.info/ The author really dives into Platforms vs Marketplaces and how to increase your chances for success. It's the only blog I've ever subscribed to.

[kromodor]

This seems pretty sound.

It should provide benefit of publishing inside regardless of the inside crowd (to a point of course).

Another example like yours - soundcloud.

[SkyMarshal]

A) Do it in Q/A form, where people can ask questions and answerers get some kind of reward/currency/rep token in exchange for taking the time to answer. That way the initial content - questions - is low cost and easy to add to the site, and the high-value content is compensated for with something other than money out of your pocket.

B) If that's not an option, and you can afford this, fake it. Go to a freelance writer jobsite and find freelance writers who specialize in bootstrapping content for social news/discussion sites. Hire a bunch of them to create the initial content..

[yaur]

StackExchange is not a great example since it was started by Joel Spolsky and Jeff Atwood who already had a significant platform where they were engaging with their target market without user generated content.

[sergiotapia]

In my case, we have a sort of the "Guide with the most views/shares get this prize X - every week!"

It's a minimal investment ($30 per prize) on our side but the yields are enormous!

[deancognation]

I was kind of wondering if we should do this for http://www.LiveFanChat.com accounts, eg king of the season at the end of a year etc.

Can you give more details about what you did?

[Gilliam]

You should know the difference between users, like readers and producer. Obviously, you need producer come to your website and create the content, if you don't want to "fake it" all the time. Readers will be attracted if the content is good enough.

So the problem is how to convince the producer to write here even though there is nearly no reader here.

Maybe "share" can be a good option. Producer can take your website as a tool and you can show the content to more users.

[anewfounder]

I have a sort of similar experience. I am developing a jamming platform and to lift it off we're looking at paying/artificially creating accounts that imitate real users. Its not cheating or anything, its just creating great content and making it look like at first this is coming from genuine users. Another method is to ask early adopters here (and possibly on reddit) to have the first go at providing content. some will respond :)

[ecesena]

Import some content from other services (under their terms).

You can promote your service to the users that they've been featured in your site and give them something better within your service, so they are willing to contribute through your site (perhaps including crosspost on the original site).

If it makes sense, you should focus on creating an initial user base with high network density, i.e. people very much connected with each others.

[sharemywin]

I started a classifieds site a while back. I advertised it on google adwords. 10 cents a click US only. it got users. I spent $1000. It cost me about $1 per post with email address no email validation. stopped advertising and traffic dropped like a rock. If your interested in doing a tech news faq site. I might be interested in working with you on it.

[lnanek2]

A founder of Reddit came and talked at a hackathon in NYC once - he said everyone involved ran tons of fake accounts at first. :)

[maxk42]

I've done this. I started with introducing it to a few friends. We made 11 posts its first night. Then I spent $15 on an ad as a sort-of "pre-launch" to see if it would stand up to a little bit of user testing. I checked back in 10 hours and the site had over 100 posts. It was a good start.

The site eventually made it to the top 200k on Alexa.

[wilzy]

Incentivize contribution by giving them something they want (recognition, exposure, points, $$$). Dangle this incentive in front of them in all the other places they currently hang out. Try to make the incentive one that also requires competition among the rest of your community.

[dlevine]

From what I've heard, Quora was originally known as the place you go to get your questions answered by Charlie Cheever. I believe that Adam D'Angelo said this at his Startup School talk, but I may be mistaken on the source.

[emhs]

I've seen a couple new sites out here in the bay try paying "beta testers" $5 or so each to contribute a little content. I remember doing that with BlockAvenue (they advertised via Craigslist SFBay).

[Theodores]

Be the expert/guru/magician/cult-leader/rock-star and grow an audience through the power of 'cult of personality'.

Let's take an automotive analogy.

Supposing you wanted to have a Q+A site for people wanting to get the most out of their Model T Ford cars, get writing and become the authority on the subject. Don't shy away from injecting a bit of personality in there and use your superior web-saviness to get better SEO, page-load times, responsive-ness and so on than what others already writing on the subject can do. A simple blog will suffice for the purposes of getting initial interest.

After a while a regular audience will materialize and they will start talking to each other in the comments. Some might even start offering spare parts to one another, uploading pictures and posting links to videos. Ideally they are drawn to your site to see what you have to say this week however they stay for the comments.

For this automotive analogy you would probably want real world community things, e.g. meet ups, galleries, a way for people to enable film-makers to hire their Model T's, a supplier's directory, a 'wiki' and so on.

Once enough 'user generated content' arrives you can move the 'forum' (or whatever) to the homepage and relegate your original articles to some archive buried somewhere in the site. By now some users might have branched out from 'Model T' to things like 'Fordson Tractors' and make the site the place for discussing everything 'Fordson'.

The need being met might be people wanting to show off their cars to others that truly appreciate such things combined with a need for people to get specialist parts. This arrived at set of needs is very different to what the site started as, e.g. the best resource on the topic. A deliberate pivot is made and the original specialism broadened.

There are many sites that have grown with a variation of the above. Cult of personality works for many writers, rock stars, diet experts, politics pundits, celebrity bloggers and probably even z-grade porn stars. Assuming you are none of those things and don't have a fan club, just by writing stuff people enjoy reading you could create a name for yourself to be someone in your field. You could become the expert in Model T Ford cars, respected for that for people that care about that particular car. Clearly in this example the market is only so big so interest might not be exactly huge, however the principle is the same. Just serve a poorly served niche with something better and grow from there.

[ColinCera]

Two excellent thinkers in this area:

http://andrewchen.co/recent/

http://platformed.info/

Read their articles.

[danbmil99]

Youtube founders have talked about this. They definitely had a "fake it until you make it" approach. I seem to recall they paid attractive women to upload videos, among other sketchy schemes.

[tiagotalbuquerq]

Fake your traction until you get your own, do the things that don't scale :) http://paulgraham.com/ds.html

[bdcravens]

I think you want to take a page out of the "Internet marketing launch" playbook and focus on growing a mailing list of those who are a fit for your website.

[bluzeee]

Gasbuddy.com is a good example too.

They pay in gas coupons for top contributors, through different means.