Digg failed because the management forgot who their audience was.
They abandoned individual user curation of links in favor of allowing
publisher accounts to feed their RSSs directly into the link submission
chain. The community didn't want to see self-serving links from Digg's
corporate partners. What made Digg great was that there were real people, scouring the web and hand-selecting links for the community's approval.

Plus,
Digg tried to chase the success of Facebook and Twitter by attempting
to emulate them (by needlessly incorporating FB & Twitter-like
features), instead of focusing on their core competency or the strength
of their community.

They inevitably scaled back on their changes
and tried to restore some of the features that appealed to their
community, but by then the communal horse had left the barn for greener,
Reddit-er pastures.

The most salient opinion I've heard about this doesn't necessarily have anything to do with Alexis or Kevin. The idea holds that the Digg userbase was essentially a boy's club when it hit critical mass, and that Digg generally didn't do anything to actively transition. Eventually, the community tired of the same, churning categories of "hot news." (And some believe that Digg was exceedingly easy to game, which probably didn't help.)

Reddit, on the other hand, has subreddits, which allows users to tailor their attention to various communities. A new reddit user can be drawn in by the friendly attention-grabbers defaulted to the front page (e.g. http://www.reddit.com/r/aww/) and then progress to discover more in-depth discussion about specific topics through subreddits. Because most subreddits are not front-page defaults, there's less incentive within those communities to capture the community's attention and clicks at the cost of quality.

Digg provides an important lesson to everyone building a startup: what succeeds in Silicon Valley will not necessarily grow beyond it. It's relatively easy to build something that will appeal to techies; much harder to grow something to massive scale.

There were a few problems with Digg:

• Spam. As its popularity rose, it became more and more important to spam it. Whenever you show the same view to everyone, spam is going to be a huge problem. Twitter and Facebook have their share of spam, but it's mitigated by the fact that everyone has their own view based on whom they choose to follow.
• Too much effort for casual users. On Twitter, if I want to share something, I just paste in a link and I'm done. Digg required me to verify someone else hadn't shared it, write my own summary, etc. Most casual users won't do this much work.
• Cliquishness. Because casual users won't do that much work, digg ended up being a small contingent of users, making it hard for others to break in. These power users ended up having too much influence over management and by catering to their needs, digg missed the bigger picture.

Digg also lost because of a fundamental shift from a broadcast model to a personal relevance model.

Digg was run by the power users basically. Digg knew that (somehow it was a flaw of their own early design) and tried to correct it, to take back the power.

With the new re-design they did just that or they thought they did (Digg didn't consult with its users about it).

Power users revolted, flooding Digg with posts from Reddit. For quite a while you could see posts on Digg which redirected to Reddit (and at that time Digg was more popular than Reddit).
Digg couldn't do anything about that (unless they would shut down the site or shut down that bunch of power users). In the end, Digg tried to reinstate some of the older design but by then it was too late. It went to a death spiral while Reddit climbed.

This wasn't the single thing which got Digg's demise but it was an important one. When you alienate your power users you can expect some retaliation.

A lot of good points made here.

I'd add only that in any social tool, bias baked into the design of core social features can produce effects at scale that are both hard to roll back and may even be toxic to culture and adoption.

This was the case, to some degree, with Digg. When it had one, its leaderboard served as a magnet for a certain kind of user and activity -- namely, promotion. Reciprocity (which is key to Twitter's follow/follow back ethic) became a norm among promotional types on Digg. They "gamed" the system -- but only because Digg enabled the game in the first place.

Culturally, the folks at the back of the room whose hands are constantly popping up in agitated grabs for attention ruin it for everyone.

Digg should have reigned in these users sooner -- but you can't fault Digg for having supported their most vocal and active users. It did net them growth (if in the wrong crowd).

In terms of collective wisdom or intelligence, I'd use that term carefully.

Collective intelligence is, IMHO, the byproduct of social interaction *if* the social actions, gestures, and system feedback loops connect to content through intelligent choices. In other words, if a system is built to capture user tastes, it will get social flavored as popularity. Not intelligence, just preference.

If a system is built to capture distinctions (hunch.com), it gets something more intelligent.

But whether this kind of outcome is "collective" is not just a matter of how many people use it. In design terms, you get "collective intelligence" only if you bracket social interaction to the side, so that the "intelligence" is a product of aggregated social activity -- not the activity of consensual or collaborative communication. Hunch works by isolating users from the social bias inherent in tastes, opinions, ratings, popularity and so on.

There was no such thing on Digg but a bipolar voting model -- yes/no. That kind of social action reinforces the Yes vote with "I agree" votes. Yes votes accrete and gain momentum. Bias is strong and on Digg was reinforced by a leaderboard that rewarded Yes vote activity.

Digg was not so much about collective intelligence as collective stupidity. That's harsh, but a vote reduces rational expression to its simplest gesture -- and as a form of communication, to its lowest common denominator.

Collective intelligence is an interesting idea -- but really may have to use non-social interaction models to succeed. Social intelligence may be an interesting alternative, if systems could be designed that had built in protections against the trending feedback loops that tend to bring them down.

Web 2.0 was not about Collective Intelligence. Digg and Reddit and Delicio.us weren't either. They certainly opened the door for crowds to let their voices and choices be known, which is good. But voting stuff up or down is a far cry from "real" Collective Intelligence. Not that we really know what that is yet. But, surely, it will be a lot more complex than simple stuff like that.

Why Twitter is winning out, is not, in my opinion, because there's anything deep and wise about a 140 character limit or about the follower model. The model is, however, flexible enough to allow for some of that real complexity which is needed. It doesn't get in the way of people connecting and self-organizing in a myriad of ways. We could say that why Twitter works better is because it isn't overly ambitious. It doesn't try to tell us how exactly good stuff should emerge from what we do. It doesn't impose any voting or rating mechanism on us. It leaves us relatively free to do the stuff we want to do, which allows more collective intelligence to emerge.

Because a following/friend model can provide each user with a personalized experience while a site with one homepage cannot. As a result, sites like Digg converge to a self-reinforcing community that is often hardcore and niche, rather than relevant and personalized to a mass audience.

However, it should be noted that the question is flawed/leading because Digg is still quite successful, just not dominating.

The site has 9M+ uniques/month in the US and is ranked #110 in the US by Quantcast: http://www.quantcast.com/digg.com. A better phrasing of the question would probably be something like, "Why did Twitter and Facebook become top 10 sites in the US whereas Digg did not."

First, I'm not sure Digg.com has failed, it just isn't as popular as it once was. But here are my guesses as to reasons.

• Digg and Diggnation are pretty unrelated; Diggnation watchers and Digg users aren't necessarily the same groups of people: some diggnation viewers use reddit, and many Digg users have never heard of Diggnation or Kevin Rose.
• Diggnation doesn't really need a super-active Digg.com for the show to work. It just needs two dudes who can bullshit about news stories.
• Digg users revolted to a site redesign, it doesn't make sense for this to affect Diggnation viewers.

In my opinion, Diggnation lost its appeal when Kevin and Alex got too rich to be relatable and spent half the episode talking about all their expensive purchases.

Digg is on death row partly because of a bad redesign, partly because the management was outgamed by the users, and partly because too many of the comments to posts are foul, stupid and insulting or fanboy childishness. Much of it seems to have been taken over by trolls.

When Digg was a poster boy for Web 2.0, its momentum kept it going, but I expect that almost everybody with any sense has already left for one of the more intelligent environments such as Reddit or Metafilter or even Twitter.

Digg is still usable as a source of links and resources, but there are now lots of alternatives. As a social space, it sucks. As a former Digg user in its early tech-oriented days, I won't be sorry to see it die.

Update: Compete's numbers for US users don't go back to 2011, when I answered this question, but Reddit now has 10x the number of unique visitors on US data. See Digital Marketing Optimization Solutions

Digg was meant to be a sharing site centered around technical question (a modern SlashDot) but, as pointed out by Soroosh, it didn’t remain true to that, and most people would compare it to the more generalist reddit, rather than Hacker News or Stack Exchange.

Failure always has many reason, but in Web 2.0, the only offense is not reacting, or not reacting appropriately to a problem your comminuty of users generally aknowledge is there. As remarkably told by Danah Boyd (ethnographer) http://www.danah.org/papers/FacebookAndPrivacy.html
Zuckerberg saved Facebook after the worst possible change in features with simple words: “We heard you. Trust us.” and asking for a few days of clemency and frenetic re-coding to correct. A few days.
Digg was riddled with very common problem:

• dilution of both:
• the theme (tech) as the site was widely featured very early on, but also thanks to Kevin Rose (business person)’s fame form his TV show,
• the quality, initially quite high as most tech entrepreneurs would step in;
• as soon as the site accepted not being so tech-centered, a handful of incredibly commited users (MrBabyMan, IIRC) who opened the door to an apparent central editorial quasi-control and revealed algorithmic cracks and possible social hacking;
  Social filters on such a large scale (often called complex, or scale-free) present odd behavioural regularities. Digg was well equipped to fight those, thanks notably to a very inventive tool to visualise stories breaking out. However those were confidential, and not enough to isolate guild-like behaviour, so the maintained status of active members lead to urban legends around of systemic gaming, cross-accusations and witch hunts;
• finally some of those ‘groups’ were suspected to do so for profit.

Those happened faster than on other sites, and Digg was one of the first site to have those as the main problem: Friendster had those too for instance, but the server crash were far more worrying, MySpace was dealing with too much friending and drama around underage sollicitation.

Digg tried to adress those, but refused confrontational or pro-active options, especially for the issue of editorial control. One option would be to cut one user or a group to have more as soon as they reach a certain share of the first page; once this is in place, a more subtle algorithmic solution could emerge, or behavioural changes (namely, more distributed attention) could prevent to concentrate that much influence. On that matter, reddit or early Quora, who learned from Digg, were far more efficient. Quora probably too much, and the visible strict demands seem to have peeled off the site’s dynamism.
Also, users were aware of what happened to Digg, and developped a keen understanding of how self-referential humour can kill a site: reddit with the idea of ‘circlejerk’, and Quora with posts like this great introduction from Xianhang Zhang:
http://blog.bumblebeelabs.com/social-software-sundays-1-humor-on-the-web-how-to-stop-it/
In that regard, I don’t think that it is an accident that Yishan Wong, Quora main contributor, became reddit’s CEO: he has the most detailed understanding of such groupthink mechanisms, how to implement techno-social solutions to the problems they cause.

Finally, to the key element of your question: no, Digg did not fail—the service, the audience, the tone of the site or the valuation are failures, however the honesty of the founders lead them to understand their mistakes. Since, by example and analysis, they have taugh others, and allow sites to overcome them. Digg ambitions and struggles are alive and well anytime a redditor sends another one a Secret Santa gift, when a bad policy gets voted out of parliament thanks to politically-savvy social media, whenever a new startup implements nuance reputation metrics.
I know the name has became of symbol of corporate abuse of those ambitions, but I remember a time when Digg was so cool it was hard to believe such a site could even work. Things haven’t change for the worst since, neither have participants; the URL might be different.

The traffic Digg used to be able to send at its peak was staggering. After some changes made to the system a few years back, they lost quite a bit of their fan base. I don't remember the exact numbers, but I think they fell from around 20m monthly unique users to 3m, and they are still falling.

Reddit never had the same massive popularity that Digg had, but has been rising steadily since Digg's redesign.

If you are looking for a place to invest effort and resources toward future audience, I'd say, without hesitation Reddit *based on the potential traffic* the site could send.

However, and this is a big caveat, Reddit and Digg are communities. They are not so huge that they have sizable enclaves that appeal to every niche. If you don't have the type of content that appeals to their users, I wouldn't bother spending any time trying to woo them.

My advice is to spend some time on those sites, see what kind of content hits. Check out if any of your own content is hitting organically, and then make a call.

Jeremy Edberg:

Any data you have to back up the claim that Reddit was twice as popular as Digg before the new version launch?

According to Alexa, right before the launch in Aug 2010 Digg was more than 2X as popular as Reddit. But right after the launch, popularity of Digg dropped like a rock while that of Reddit shot up. It's pretty clear in this graph:

Absolutely, I’d love to share my experience with using Digg and Reddit as social media tools and how SuSocialPro transformed my approach. Spoiler: It’s been a game-changer!

Once upon a time, I was drowning in a sea of social media accounts. Managing my presence on platforms like Reddit and Digg felt like trying to juggle flaming torches—blindfolded. I was constantly switching between tabs, trying to remember which account needed what post, and my engagement? Let’s just say it was less than stellar.

Enter SuSocialPro, my social media knight in shining armor. This AI-powered tool took the chaos of multi-account management and turned it into a symphony of streamlined efficiency. Seriously, it was like having a personal assistant who never sleeps!

Challenge #1: Multi-Account Mayhem

Before SuSocialPro, my browser was a mess of open tabs, each logged into a different account. I felt like I was on a never-ending carousel of confusion. But with SuSocialPro’s multi-account management feature, I could finally breathe. It allowed me to oversee all my accounts from a single dashboard, making cross-platform operations not just manageable, but downright enjoyable.

Challenge #2: Engagement Woes

Engaging with users on Reddit and Digg was another beast entirely. I’d spend hours manually posting, commenting, and liking, yet the results were lukewarm. SuSocialPro's automation tools were like a magic wand for my engagement strategy. With automated posting and messaging, my presence became more consistent, and my interactions, more meaningful. My engagement rate skyrocketed, and I even had time to enjoy my morning coffee!

Challenge #3: Understanding My Audience

I was also missing the mark on understanding my audience. I felt like I was always speaking into the void, unsure if anyone was listening. SuSocialPro’s user analysis and segmentation features changed the game. The AI-driven insights provided a crystal-clear picture of my audience, helping me tailor my content to truly resonate with them. It was like finally finding the missing piece of the puzzle.

Surprising Solution: AI-Generated Content

One unexpected delight was SuSocialPro’s AI-generated content feature. It helped me craft engaging posts that were tailored to each platform’s unique vibe. The time I saved was immense, and my content quality noticeably improved. Who would’ve thought an AI could become my creative partner?

And here’s the cherry on top—SuSocialPro didn’t just enhance my Reddit and Digg game; it transformed my entire social media strategy across Instagram, Facebook, and beyond. If you’re in the same boat I was, juggling too many accounts with too little time, I highly recommend giving it a try.

To sweeten the deal, use my exclusive discount code JA45%OFF to get a 45% discount. And for more practical insights and tips like this, feel free to visit my Quora homepage.

Trust me, SuSocialPro is like having a superpower in the world of social media marketing. Give it a whirl and watch your digital presence soar!

Probably no one thing.

Here's the beautiful front page of Digg:

Lots of interesting content, presented beautifully and simply.

Here's the front page of reddit most people see:

Here's my front page:

Aside from the silly moose - you'll see that a lot of the default subreddits are gone. I don't enjoy them as much as I used to and I've shifted to spending a lot more time in the smaller communities that I'm more interested in (homebrewing, Android development, startups). All these little communities power a large portion of the user engagement of reddit. They act essentially like forums. The most likely contender for this portion of reddit is probably some thing like Discourse.

Take a look at this blog entry:
WW1 Books, LennyBot, Zombie Jesus Pizza, Confronting Procrastination, and a Good Roux: the most saved comments and posts on reddit.
It describes the most saved comments and posts on reddit. A large portion of them are on /r/AskReddit. If one was to guess at a reasonable competitor for that, it would have to be Quora - but Quora's framework is much better suited for long single-answer replies rather than short multiple answer replies. Reddit can churn out quality responses at scale in a way that few others can.

Now, I don't want to post pictures of 4chan because that's never a good idea, but the last portion of reddit that sits somewhere between Digg and 4chan is the social media aggregator. The big community of reddit is fantastic at promoting its own internal culture and jokes. The other major contender for something like this is 4chan.

If one site wants to replace reddit it has to be able to provide these features right off the bat AND more. Reddit doesn't succeed because of its HTML and CSS, it succeeds because of its community.

All this said, reddit only gained the prominence it did because Digg managed to upset its users enough to cause a mass exodus. Reddit will have to do something like ban all mention of bacon, your girlfriend, Mario and Portal in order to incite insurrection and have everyone flock to a new site. Maybe that's a good thing.