Why Google *really* killed Reader

[drwex]

http://www.marco.org/2013/07/03/lockdown

Marco Arment has a pretty cogent analysis, which boils down to "because Facebook."

Worth reading.

Comments

[dr-memory.livejournal.com]

Rehoisting a comment I made over at jwz.org:

The baleful influence of the Plusterfuck was, sadly, only the final nail in Reader's coffin.

For the most part, Reader was killed by Google's internal corporate culture, which prioritizes and rewards enormous ocean-boiling projects over thoughtful stewardship and incremental improvements to mature products save for the (mostly) untouchable cash cows of search and ads. Nobody wanted to work on Reader because working on Reader was clear and obvious career suicide: you would get no recognition from your peers, no shout-outs at TGIF, no multiplied bonus, no big lucite award plaques... and in the end no promotions and no job.

Reader was and is far from alone in this category, and all that was the case long before Vic and Larry ascended to the throne and made FriendsterFacebook into their white whale-- just ask the Dodgeball guys, or the XMPP team, or anyone who still admits to working on Google Voice.

And... that is what it is, I guess; they have and occasionally continue to put out useful products despite all that. But a product at google with no engineering support is a dead man walking, because it's inextricably built on top of a set of infrastructure products (bigtable, GFS, etc) that are themselves constantly changing, with no guarantee whatsoever of backward compatibility between releases. Eventually one of the shared libraries your code depends on gets removed from the relevant package repo, and you can't launch any more servers and you're gone.

[drwex]

Fascinating. Thanks for the insight.

[sandhawke.livejournal.com]

Normally in an industry when there's one runaway monopolist leader, everyone else teams up to take it down, without a lot of infighting.

So it would make a lot of sense for Google to support standards in this area, getting everyone to team up against facebook.

Wouldn't it?

cf http://www.w3.org/2013/socialweb/

[drwex]

I believe the linked article explains (gives the author's view on) why companies are running away from standards at this time.

[sandhawke.livejournal.com]

Yes, but it seems implausible to me. I mean, it's basically saying that Google (and everyone else) would rather try to beat Facebook at its own game, in a winner-take-all match, instead of changing the game to one where everyone shares the pot. Since Facebook is clearly, hands down, 100-miles-out-in-front in it's own game, this basically makes everyone else either evil or stupid.

[drwex]

Well, you are of course welcome to your own opinion but I don't see it as stupid. It's a continuation of the strategy that has pitted large tech companies against each other since approximately forever.

Abstractly I think it resembles a non-iterative prisoner's dilemma in that companies believe they will gain more by competing than cooperating. In particular if you believe that non-cooperation would remove your opponent then it's logical not to cooperate. See AOL and Yahoo who still exist, technically, but who have been effectively removed from play, along with hundreds of other smaller players. What's interesting is that smaller players who have no hope of knocking out the giants are playing non-cooperative. But if you are in a prisoner's dilemma - iterated or not - and you believe your opponent will not cooperate then the logical strategy is for you not to cooperate either.

[sandhawke.livejournal.com]

Sorry if my language was offensive.

But we have lots of evidence of cooperation, with wild success. The list of standards on which these companies is built is vast and very well known: Ethernet, USB, SCSI, C, C++, Java, IEEE floating point, IP, TCP, HTTP, HTML, CSS, JPEG, XML, JSON, SQL, ASCII, Unicode ... just to name some outrageously successful standards). Without the cooperation that produced these standards, none of these companies could exist. Most large and many medium-size tech companies companies put lots of work into standardization, paying consortium dues and allocating multiple FTEs to these efforts. Even Facebook is now quite active in parts of W3C.

But it seems like when it comes to social, they forget all this.

[drwex]

I'm not offended.

All the standards you cite either were developed in days when companies saw the competitive landscape differently and/or were international body efforts and/or did not appear to allow others access to what companies saw as their proprietary grounds. When they did (e.g. Microsoft vs Java) they fought hard against.

As I think the original article argues, Apple and then Facebook changed the game. Apple has made a religion of doing things its own way, closing off the rest of the world, and being ridiculously successful at it. AOL failed at this game, but Facebook succeeded using essentially the same playbook as AOL tried. if you have two models in front of you: closed/proprietary/wildly successful or open/sharing/um... yeah, sort of less successful which model would you emulate? That's rhetorical - I'm just saying it's not an irrational move to emulate the big successes.

[sandhawke.livejournal.com]

But it's not a question of repeating Facebook's domination -- it's a question of ending it. The same game is still running, and Facebook is winning handily. There's one well-known tactic for bringing down a vendor that's utterly dominating the market: team up against it. Facebook can beat Microsoft OR Google OR Apple OR Twitter OR IBM OR LinkedIn, but can it beat all of them working together against it? Not so likely. This is what everyone did with Java and Web Technologies against Microsoft, roughly speaking.

It seems like what folks are doing now is like if Sun had kept Java closed to the degree Microsoft would have, and tried to compete solo, 1-1 against Microsoft. Or everyone had stuck to their own propriety networking solution (think Novell and IBM Token Ring) instead of using Internet protocols. That would have been foolish of them, yes?

[drwex]

I admire your altruistic sense, but I don't think it's widely shared in a business context, which is seen as competitive first and cooperative only when necessary.

The question of whether Facebook could stand against a unified alliance is moot because such an alliance would be unstable at best, illegal under many circumstances, and unlikely ever to form. The business cost of fighting a group of competitors has to be weighed against the business costs of lowering barriers to entry (which is one thing standards do) and then allowing lots of other people in to compete.

I don't think Facebook sees Google or Microsoft as its biggest threats. Their real threat is going to be some company nobody has yet heard much about who will do to Facebook what Facebook did to MySpace and their ilk. To the extent that Facebook exposes APIs anyone can use, and subscribes to standards anyone can copy it's enabling that anonymous future entity to kill it faster and more easily.

And again, remember that your original supposition was that acting in a proprietary manner was not "plausible." I think this strategy is perfectly plausible. It may turn out in the long run to be inferior to your preferred cooperative strategy, but I don't see how you can continue to maintain its implausibility.

[ringrose.livejournal.com]

What I don't understand is why Google failed to incorporate reader into g+. Turn each of your feeds into a line of posts exactly like your friends' posts which you can include or exclude, and then gradually start pushing people from using reader to using g+.

[drwex]

It's my experience that Google doesn't do much of anything "gradually." One day you wake up and you're using Plus in email and your Picasa albums are all gone. One update on your Android phone and you no longer have chat, you have "hangouts". Sometimes they allow you to hang onto "no, don't take all this away yet!" options. Google doc did that for a while, but now it's all Drive whether you want it or not.

[dr-memory.livejournal.com]

They actually tried: they replaced Reader's internal sharing system with g+ circles. The result was an immediate loss of something like 15% of Reader's active users, a huge outcry from many of the remaining ones (who liked the existing system just fine, and like all sane people had no urge to join the Plusterfuck), and no increase in sharing usage by the remaining.

Google does not grok "mature product with a stable and happy userbase."