Lessons from the failure of Livejournal: when not to listen to your users | Hacker News Reader

Lessons from the failure of Livejournal: when not to listen to your users

arvindn.livejournal.com

72 pointsrandomwalker24 comments

Always listen to your users. Don't always do what they ask.

When I took a course on software usability many years ago the first principle was how to work from what a user asked for to what they wanted to do. So, a user might say, "Can I have a checkbox to turn off feature Y?" What that meant was, "The way feature Y is implemented annoys me." And the solution is often to rethink feature Y and what it's supposed to do.

Exactly. Henry Ford heard "I want a faster horse" but gave people an automobile. Because he heard "I want to go faster."

To be clear, Henry Ford was a great innovator in production and assembly techniques, but Karl Benz created the first automobile 23 years before the Model T. I suspect what Ford heard was, "I want an automobile." He figured out how to do that cheaply and on a large scale.

Karl Benz created the first automobile

The first automobile was created before Karl Benz was born.

http://en.wikipedia.org/wiki/History_of_the_automobile

Benz created the first production automobile in 1888, 3 years after his (and, respectively, Daimler's and Maybach's) first automobile, and 20 years before the first production Model T was built at the Piquette Plant on September 27, 1908.

oh my god, you guys. way to miss the point.

No, stop. It is important to be able to object precisely and factually. It is trivial to look up information, and leaving misinformation in a comment because you can't be bothered to check the Wikipedia articel is as jarring as that typo I didn't bother to correct a few words ago.

I don't disagree with the statement that Henry Ford gave the public the automobile (instead of merely a better horse). I agree with your analogy (as long as "faster" can be replaced with "better" -- since horse-based transport had other problems besides merely being slow). My understanding is that all other efforts, around that time, toward automobile production fucused on charging as much as possible per unit, and limiting production toward that end. Even Ford's early investors felt that that was the only profitable way forward, and he had to fight to get them to agree to do the opposite -- minimize price and maximize production.

Henry Ford, more than any other individual, replaced America's, and the world's, horses with the automobile. (I use Ford's 1922 book My Life and Work as reference for these statements. http://www.google.com/search?q=ford+my+life+and+work)

And while we are on the subject of analogies, we might mention Sony which, in 1983, instead of giving the world a better analog phonograph player, gave the world the Compact Disc player.

So true. Probably the first lesson I learnt as a product manager was how to take what users asked for in terms of features and then try and understand "why" they wanted it. Especially useful when you listen to a number of feature requests and try and distill them down to the reasons underlying them. Many disparate ones can often be coalesced down to a few actual requirements/issues which you can then try and address instead of just implementing features

LiveJournal isn't a failure! It has millions of users! It is responsible for substantial open source projects! It pioneered social media! It made a few people rich!

I don't drive a Model T, but a lot of people did.

Show some respect.

Hear, hear.

Memcached is damn useful, and Perlbal saved me from server meltdown and banning many a time when it was still my favorite language to write the hackish, duct-taped monstrosities I called "web applications."

LiveJournal hasn't been the same since bradfitz stopped working on it, but it was a milestone as far as interactivity on the Internet is concerned.

This post misses the main reason LiveJournal is a niche site today: they were invite-only for years. LJ had severe performance issues during the first couple of years, and to keep the system usable during that period new accounts could only be created if you procured an invite from an existing user. By the time they fixed the problems and removed the invite system, LiveJournal had been eclipsed by the growth of other social-networking sites.

They opened up to the public in 2003, which is when I signed on. They had memcached by then. My analysis is about the last 5 years. Other social networking sites did exist, but they didn't let you blog or read feeds. LJ had something unique.

Unfortunately, MySpace started up about six months before LiveJournal removed the invite system, and clobbered LJ by seeding themselves with 20 million accounts on the first day. At that time the demographics for MySpace and LJ were roughly the same, and most people simply followed their friends onto MySpace. The coup de grace came a month after the invite system was removed, when Facebook was launched.

I'm not saying that the invite system was the only mistake they made, but it cost a year or two of lead time and allowed competitors with much deeper pockets to make up the gap.

Its interesting that there's no mention of SixApart. My theory would be that SA bought them to kill them.

Possible over simplification, but powerfully relevant statement right there. SA bought LiveJournal and the site somehow managed to get worse. There's no need to implement an ad infrastructure to the site yet that's exactly what happened.

I used to use LiveJournal, as did a number of people that I know and whilst it was never perfect, we all agree that it experienced a significant decline once it was bought by SixApart.

I would guess the reason as being Brad's phaseout (as a result of SA acquisition) more than a straight conspiracy theory. It was pretty odd that they launched Vox shortly after buying LJ, which was basically the same thing.

Yea but LJ just felt more nostalgic, and just as the article said the community feel of the site was just great.

Does that really make any sense? Once they owned LJ any success it had was theirs. "Never attribute to malice that which can be adequately explained by stupidity."

Companies are made up as individuals, and the actions of a large company are often not just the actions of the top management. And I didn't mean to suggest malice rather than stupidity, because it could have been entirely due to "institutional" stupidity.

It's interesting to read this, since I was a member of the fandom contingent of LJ from 2002 to about 2005. I guess that makes me a social deviant. ;-)

Anyway, there's a grain of truth to this. The fandom community is disproportionately early adopters. After all, people with mainstream interests can hang out with their friends in real life instead of meeting them online. Many of my fandom friends were among the first users of GMail (I got my invite from one of them), they were among the first to switch to FaceBook when it opened up to non-college students, and now they're the ones that are on Twitter and checking out FriendFeed.

However, I think there were many, many other problems with LiveJournal that prevented it from gaining mainstream adoption:

1.) Until about 2005-2006, the front page never really explained what LiveJournal does or why you should have one. If you didn't have friends that were already on LiveJournal, you didn't know how to use it.

2.) It was plagued by perpetual scaling problems - in 02-04, it was virtually impossible to get your friends page to display. It was as bad as Friendster, and worse than nearly every other social networking site.

3.) Their decision to open-source their code meant that whenever they did anything controversial, someone would set up a competing site and a large fraction of their userbase would migrate, coming back when that site went out of business. When they had their scaling woes and required invite codes to get a new account, everyone went to DeadJournal. When they lowered the number of free icons, everyone went to GreatestJournal for icon hosting. When they pissed off the dev community by calling them ("Nobody but me could run this site" -- Brad Fitzpatrick), Plogs.net got started. When they banned fandom_wank, it just migrated to Journalfen.net. It's really not a brilliant idea to charge money for your service and then give away your source code for free.

4.) They were just too early. Same reason that YouTube, Comcast, and Verizon Fios got rich providing the same service that Enron Broadband had wasted billions on. Mainstream adopters were not ready to shift their social lives online in 2002.

In particular, the article suggests that if they'd let you import external RSS feeds into your LJ friends page, they could've become the default feedreader for the web. There're good reasons to believe this isn't true, notably:

a.) This describes Bloglines, which came out in 2003 and wasn't notably more successful.

b.) It does not describe FaceBook or MySpace, which were successful. Really, the whole "feed" concept seems to be a bust among mainstream users - PointCast failed in Web 1.0, RSS never caught on outside the technorati, and FriendFeed is early-adopter only.

c.) They did let you import external RSS feeds into your Friends page - it's just that free users were limited to no more than 3 feeds (paid users still had a limit, but it was higher), making it not so useful.

The article also mentions pingback support - my friend implemented trackback support for LiveJournal in 2003, and it sat in the CVS repository for 2 years without ever being merged into the production codebase.

Really, if I had to pick a reason LJ went nowhere, it's simply that it was too early. And it's not like it was a super-huge failure too: the founder still got his f-u money.

Didn't Livejournal create Memcached?

Memcached was created by Danga Interactive, the company which ran LiveJournal, to improve LJ's performance. See http://www.danga.com/memcached/ and http://en.wikipedia.org/wiki/Memcached .

You're not the only one to mention memcached. Please go read the article, that's part of what I'm complaining about. They had memcached, and so many other things, including a very advanced commenting system, and yet they lost the opportunity to scale and go mainstream, and remained a niche site which consists largely of support groups and fanfiction. (Again, nothing wrong with that, but it could have been so much more.)

Reply on news.ycombinator.com