Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
If I were starting out now (in startup land), I would clone threewords.me (gabrielweinberg.com)
59 points by tchae on Feb 2, 2011 | hide | past | favorite | 32 comments


Kinda huge assumption that you are going to go viral. Mark himself said in some earlier thread that he has been trying many things for many years and this was the one that took off.

Bottom-line is that you can predict what would go viral as much as you can predict what would be fashionable. Not that much.


On the other hand, you can predict with 100% accuracy that any idea you _don't_ build _isn't_ going to go viral. (at least not for you...)


Ahh the dangers of 100%... A notable exception: Tyler and Cameron Winklevoss ( http://news.ycombinator.com/item?id=2054297 )


Maybe... I suggest that having an idea, not implementing it, then suing the guy who did build it (several times) doesn't count as "going viral for you"...

But yeah, it can still pay out pretty handsomely...


You don't have to predict. That's why you clone :)


Some things seem like they're a one-shot though, e.g. the Million Dollar Home Page.


But this doesn't seem like one of them IMHO.


The key to something going viral is its novelty. If something has already been done before then it's almost guaranteed not to go viral.


You're missing the key of threewords.me. It went viral in a small subset of users on FB/Twitter. It could just spread similarly in many other areas of FB/Twitter, you just have to seed it appropriately.


Can you give one example of this? I can't seem to think of any. There's only one threewords.me, and there's only one chatroulette. The followers most likely won't go viral.


threewords.me != chatroulette IMHO.

The latter went viral on the back of press and wasn't inherently viral. threewords.me is inherently viral and has only spread in a small area of the big social graph. That's why it is cloneable right now.


There are many things that have only spread in a small area of the big social graph, but that doesn't mean that clones will go viral.

As stated above I can't think of one single instance where a clone product or derivative of a viral product or idea has itself gone viral. Whether that product is threewords.me, chatroulette, all your base are belong to us or hotmail.


May I suggest that this can go viral in several years time again? Perhaps as a remixed site of some sort? I mean its not necessary that something can't go viral again, given enough time to forget about it.

Analogy used here is to forwarded emails - every so often you find the same theme pop up again.

Is there any link or any comparison that can be made to going viral and urban legends? Urban legends end up tapping into human nature/brain wet-wiring and hence have huge longevity.

Perhaps sites which go viral have something similar?


Normal people can't. Somehow the real visionaries can. Just saying. In technology, Steve Jobs would probably rank at the top.


Good idea. I'm on it.


Are you for real?


I'm guessing you caught that sarcasm, but for anybody who is unaware - he's the one who made threewords.me


Oh, I caught it, but it is still a valid question :)


Hehe, nah. I want to try other viral experiments, though. That I'm actually on :)


Good to hear, but I am curious.

When threewords went viral, you were using a single linode, and seemed to solve your scaling problems by just upgrading to a larger node and implementing a bit of caching. What have you learned and what would you do differently this time?


First, it's fun. I can't argue that that may be the case.

Second, you'll learn a ton by working with millions of users. This is predicated on the huge assumption that your viral platform attracts millions of users, and that a large number of these (assuming every new internet entrepreneur is doing one...) are sustainable.

Third, it will pay off. Mark's pay-off, I'm sure, was more than he initially envisioned. That amount, to my knowledge, however, remains undisclosed. I don't believe enough evidence exists to assure such a payoff and create such confidence.

Fourth, you can use the user base to launch another startup. Sure, 1% of ~10M users is cool. But that, again, rests on the assumption that you can build up that base of 10M users.

I realize that this formula may have worked, and that value could likely be created from cloning it. Cloning just may work. I can't shake off the feeling, though, that a bit of misinformation is being spread here.

[edit for clarification]: Of course, both the author's post and my response is solely opinion, which makes my use of the word misinformation, well, misinformation. What I mean to say is that I can't help but draw comparisons to "get rich quick" schemes, where opinion was packaged in such a way that people took it as fact.


"a bit of misinformation is being spread here"

It's my opinion on my personal blog, not fact. I hope that was clear.


> Sure, 1% of ~10M users is cool. But that, again, rests on the assumption that you can build up that base of 10M users.

Also, he forgets that there are numbers less than 1%.


1% wasn't chosen out of a hat. I see it in real (startup) life a lot.


1% of 10M users isn't cool. You know what's cool? 1% of a billion users :P


Somewhat related question:

If a viral project takes off and a month later 95% of your traffic has died off, but you find yourself with 10 million email addresses (signups), is it ethical or even legal to promote a completely unrelated project by emailing everyone?

I'm sure you could word it in a way that doesn't seem too spammy ("Thanks for checking out xxxx, try our new project: yyyyy").

I'm in this situation.. I've held off so far, as tempting as it is, because I hate spam myself..


Can you update your original project with something worth mailing people about, then include in its footer and in the footer of the mail-out, something along the lines of:

"The Makeee network includes OriginalProject.com and the all new NewProject.com (has to be seen to be believed!)."

Or:

"PS. If you've got time, please check out my latest project - I'd really appreciate your feedback."

Or, perhaps a little spammy, do a joint mail-out for two projects in one newsletter?


If you don't send regular emails, don't email them all yet. With 95% of the traffic dying down, scale back to minimise costs and work on new project until it's finished/ready.

When you get a new project ready; say something along the lines of "This will be the last email from viral-project.

The viral-project closing down, thanks for participating! (Allow option for people to get messages/pm/etc if it's a permanent shut down).

I will be focusing my efforts on a new direction: new-project."


it's fine to do that in my book...hell it's not like you are selling the email addresses like some people or spamming them with Amazon affiliate links.


1) although you can hope for it, you can't guarantee something will become viral. threewords.me could have just as easily been another "project" had it been pushed out at the wrong time, sent to the wrong people, etc. etc. there's a lot of luck involved. i'm glad it took off the way it did. Mark made sure he would maximize the viral potential by having it easily shareable, etc, but that alone didn't make it viral.

2) i'm just jealous that he got real time experience in learning to scale his servers to support the crazy growth in users. not many people in the world have experienced viral growth with their webapps or startups. so congrats to Mark.


it's on now, baby: fourwords.me

(fivewords.me, sixwordsme.com.au, etc.)


Cool, and then when the enumerable infinity of naturals is registered, let's go transcendental.




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

Search: