Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Jump starting techniques for a new site?
44 points by Lorin on July 31, 2010 | hide | past | web | favorite | 25 comments
In #startups (irc.freenode.net), Phil_H brought up the interesting question of alternatives to founder sock-puppeting to jump start a website and/or community.

What are your suggestions or past experiences with this kind of situation?




The textbook answer is to start in a niche. What you need is a large proportion of the market. Aim at a small market and you can be a big fish in a small pond quickly. Then you can expand from there.

Some examples:

Gmail: Google geeks -> well-connected geeks -> all geeks -> non-geeks

Facebook: Harvard -> Ivy League -> all US universities -> etc.

Stack Overflow: Joel's blog readers -> programmers at large.

A good proportion of success stories contain a similar history, and for good reason. It is particularly easy to see why this is so effective by considering Facebook.


Organic SEO via scalable content creation on fairly uncompetitive tail search terms, with profits and exposure being used to target progressively more competitive terms and fuel more expensive customer acquisition strategies like AdWords.


I could probably scour your backposts to find this, but where do you draw the line between worthwhile keywords and completely uncommon keywords?


That depends on cost and value. For me, I can be profitable after an average of twelve clicks per page, lifetime. It is virtally impossible to think of a bingo activity too obscure for me.

Anyhow, guess and check. If you write and cannot rank for competiton, go down tail. If you write and cannot profitably exploit traffic trickle, move upwards.

Note that entire portfolio of content gets better over time if you do it right. (Superlinear returns on linear cost.)


If your site has a "chicken and egg" problem, that is it exponentially more useful the more people are using it (and vice versa) then you're going have a bumpy road ahead.

Chicken and egg suggestions:

- Do all you can to make your site not appear a ghost town. Create content yourself, and hide pages that don't function without enough content until there is enough content.

- Feature the best and most relevant content early on. This rewards your users, and gives visitors something to look at (while enticing them to join).

- Kiss your early adopter's asses; make it easy for them to give feedback and more receive personalized responses.

- Make it as easy as possible to join. This is by far the largest barrier to entry. If you can get users invested in joining before they join, that's great. For example, allow them to submit a comment, then say "this comment will be submitted when you sign up". Cheeky, but effective.

- IMO, don't show "streams" of stuff unless that stuff is really awesome. If I see a stream of stuff and it looks lame, I won't be impressed that it occurred 3 seconds ago.

- Make it look like a lot is going on. "A lot" can mean only a handful of things, which could all be made yourself.


Blog about... anything! (Well, not quite...)

The KSplice guys have been doing it, and doing it well. Eg. their traceroute tutorial. It has little to do with their product, but it does well to drives traffic to their site, and raise product awareness. http://news.ycombinator.com/item?id=1557877

(See also: Joel Spolsky)


Build an email list on the relevant subject with good information, then when you're ready to launch mail all the list members so they'll arrive in a relatively short period of time. Figure a 15% response rate (fairly normal), a few tens of thousands of people on your list should do the trick.

After that fan the flames like mad to stop the fire from going out.


I'd recommend to the book Made to Stick by Chip and Dan Heath here, despite it being more geared towards marketers, than website launches. Derek Sivers has a nice set of notes on it over at http://sivers.org/book/MadeToStick



Relevant link posted in channel with lots of methods listed: http://www.gabrielweinberg.com/blog/2010/04/in-the-pursuit-o...


To clarify my thoughts when I brought it up, it seems like the canonical example of this problem is a dating site: sockpuppeting seems somewhat unethical, and somewhat impractical (can you flirt with the guys who message you, but manage to tactfully reject suggestions of meeting? Do you have multiple sockpuppets per state? etc.), but without a seed the site is worthless.

I guess it would help that when someone signs up, they don't need to come back for people to send them messages. So maybe you don't need a community like reddit did, just users. But I don't know.

Does anyone know how sites like OKCupid solved this?


I noticed that when they (the OKCupid guys) launched Crazy Blind Date, they limited it to a couple of areas where OKCupid had a lot of users already (Austin, and San Francisco, I think were the earliest launch cities). Of course, they already have a user base...but the idea is perhaps useful, anyway. You can't possibly afford to blanket the whole country with ads, but you might be able to reach a decent portion of young San Franciscans via some reasonably priced method.


They used sockpuppets and got rid of them once they had a high enough concentration in certain areas and it could stand on its own. PlentyOfFish did that too.


Is that really true? Out of curiosity, how do you know this?


There may be others, but a class action lawsuit was brought againt Yahoo! Personals for their use of sockpuppets in converting free to paid accounts.

http://plentyoffish.wordpress.com/2007/08/30/yahoo-class-act...


On a related note, on my site, I am considering implementing an "import your profile" feature that will import a person's public domain profile information from a competitor's site. My hypothesis is that this will reduce the barrier to signup for at least some of my users.

My question is - is this legal? The data(profile) is publicly accessible and belongs to the person(technically I understand that the site owns it). However if it's out in the open, what are the terms for this?


Every site is different.

Many sites ban scraping and will only allow you to use the API. Even using the API they may only allow you to hold on to the data for a limited amount of time.


Looks like I need to kick off the discussion myself :) One concept discussed was "the velvet rope" technique where only those with personal invitations are allowed to enter or sign up.

This technique was highly successful during the original Gmail Beta launch :)


They also had the element of scarcity, not only did you require an invite, but those invites themselves were very hard to come by. USB's were traded for them, and sites dedicated to getting them, gmailswap.com anyone?


Invite-only has been pretty successful at Quora (now public) and Forrst so far. A lot of designers also want in on Dribbble because of the scarcity of invites they've created.


Google does not do this as a technique of attracting interest, they do it so they can scale more predictably. If it increases interest, it is only because people already wanted it.


Only because Google did it.


Do something not listed here :)


1. Pay people to create the initial content.

2. Create it yourself using multiple accounts.

3. Have a following like the StackOverflow guy did.

I've had success with growing it one user at a time, starting with yourself, friends and family. If your startup provides something of value, it will grow.

If it's not growing at all, even a couple users a day, then that hints towards a problem with your product/service/site and you need to get some feedback from your users. Some of your assumptions when creating/designing your site may be off.


Agreed. Another relevant link: http://www.startbreakingfree.com/1322/early-adopters-5-ways-...

Summary:

1. StumbleUpon

2. Run ads saying "[YOUR COMPETITION] sucks"

3. Long tail SEO

4. Turk it

5. Build a blog (super powerful but takes a long time to build up this asset)




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: