I never knew there were people out there who thought Reddit copied Digg. In fact, I'm not sure anyone can copy anyone in this case. It isn't like having people vote a news item up or down is a revolutionary idea in and of itself. What makes Digg, Reddit and yes Hacker News special is the community that was built around them. That, imho, is the innovation in these sites. Founders creating an enviornment in which a vibrant community can flourish.
In a previous Mixergy interview, someone with inside knowledge suggested that Reddit copied Digg. It's been suggested in other places too, so we cleared it up.
But this interview goes beyond that question. I wanted to learn where the seed of an idea comes from and how it grows from that tiny thought into a real business.
Articles being voted up/down is kind of a revolutionary idea, in that it could have been developed years earlier, but it wasn't. Why not? (maybe people tried it, it didn't take off and so we didn't hear about it)
One factor might have been ajax. Being able to vote without having to reload the page is a big deal, but it's only relatively recently that the techniques to do that have become common knowledge.
Not that that stopped slashdot from doing it with comments. It could probably have been made to work with articles as well; quite possibly everyone just thought editors were better at selecting interesting links to publish. The innovation was social rather than technical. (HN's system of articles selected primarily by voting, but with more than cursory editorial oversight, seems to be another social innovation. Did any earlier site do the same thing?)
I worked out how to do this with an invisible 0-size frame, over 10 years ago. From another frame, you put data in the invisible frame (in a form, or as an URL) and submit it; then read the data back. Of course, it was awkward, and maybe there were some cross-browser issues (I don't recall, but such issues were the norm).
But here's my point: I thought it was technically cool, but I didn't do anything with it. Not even a demo. So this is a striking example for me that it really matters how familiar you are with a concept, before you can start using it as a basis of inventing something else. Progress takes a long time, and proceeds in steps. There's a tantalizing possibility that, maybe, knowing this could give you a simple way to leapfrog a decade ahead...
Actually kuro5hin.com did the same thing several years before.
I have said in the past that reddit's early marketing blurbs sure seemed similar to digg's. But I never really meant it to be taken as a substantial criticism. Everyone copies (and more importantly modifies) other people's ideas all the time.
There's no way reddit copied digg. Sure it's feasible that they were inspired by the idea, but the sites are fundamentally quite different. Digg.com is just a chronological feed of stories that have fulfilled the front-page promotion requirements. Reddit, on the other hand, is an actual ranking of stories, so an extremely popular story stays on the front page longer and gets higher up on the page.
Paul Graham, with all his knowledge of the industry, didn’t know that Digg was out there either?
It sounds kind of weird to me that he wouldn't... and yet, for my startup, I also wasn't aware of some very similar projects (and a good thing too, or maybe I wouldn't have started)
The interviewers disbelief that food money is so important reminds me of my first experience with startups. When I was an undergrad I worked for a doomed startup, and I remember with all the various government and other sources of support out there, absolutely noone was willing to pay for rent or food. I mean the money had to specifically be for hardware or the like, but the biggest hurdle is the day to day money. Even living with your parents you need to buy some of your own food. It's not in the least bit trivial.
We'd become ramen profitable from our licensing work with Conde Nast before the acquisition and had enough to live (modestly) on for another year.
We were slow to put ads on reddit, putting FederatedMedia on only the comments pages, and generating a few thousand a month from just those pageviews (a small% of our traffic).
That said, if that deal had fallen through, we'd have likely just powered on. I can't speculate whether or not we'd have gotten acquired by someone else, but I'd been discussing the idea with a couple other properties.
Maybe we would have gotten serious about selling merch sooner ;)
For reasons I hinted at in the interview, it really was the right time for an acquisition.
We built lipstick.com (now http://weheartgossip.com) for them. I received an email one day from their head of biz dev, he'd heard about reddit through a mutual friend. I called him that day and we spoke for a good while about an idea he had for a celebrity gossip version of reddit. I told him it sounded like a fun project to work on and things went from there.
We've never spoken publicly about the terms of the contract we signed with them, but it put us in good ramen-eating shape for the next year or so. In hindsight, this great deal was likely part of the wooing process. But it worked :)
I wasn't asked. But Steve and I had a bunch of usernames we'd log into for the first few weeks to submit cool links. I must have had about 10 that I'd cycle through (such is the life of the non-programming cofounder in the first month of a social news site).
The day when neither of us had to submit anything was a great one - it happened during that summer and gave us hope that the damn thing might actually work.
Are you saying that's a bad thing? Getting traction is extremely challenging. Most startups reliant on UGC do this, in the past we've done it with Amazon MT too.
I don't think it's something to look at negatively when you take into consideration the judgmental nature of a consumer. Most consumers don't know or care how new a product is, if it looks inactive they won't use it. Faux-ing a small amount of activity at the start gives you a chance to set a tone and direction for user-generated content, and helps overcome the initial bias of a new site.
Writing your WebApp in Lisp or another trendy languages like Scala, Clojure, Erlang, etc. is a nice story for viral marketing, because the communities around those languages are desperatly looking for "killer applications" and "success stories". That's a lesson i learned from reddit.
True, but this only works if you then switch to Python and get all the Lisp/Scala/Clojure/Erlang programmers' knickers in a bunch. (I'm not English, by the way; it's just that the Brits have a gift for odd metaphors.)
