Google changes Buzz to be opt-in and adds easy delete button | Hacker News Reader

Google changes Buzz to be opt-in and adds easy delete button

gmailblog.blogspot.com

106 pointsspicyj48 comments

I honestly love Buzz, but the rollout was terribly executed. It's rather ironic, since the product itself is one of the better designed social networking tools I've seen. Like Beacon was a real defining moment for Facebook, though, I think that the monumental amount of criticism that Google has taken over the Buzz rollout will teach them to be more careful about the second-order privacy implications from now on. It's just a shame that it took this much damage for them to learn.

Can you articulate why you love Buzz?

For one, the use of open standards promises wide interoperability in the future. Right now, my first six tabs are Gmail, FriendFeed, Facebook, Twitter, Voice and Wave. I'd love it if those were somehow tied such that I could look in one place for my news, but if doing so didn't tie me to a monopoly.

Second, I love Buzz for the same reasons I love FriendFeed. It aggregates my interactions, which goes a long way towards healing the massive fragmentation problem that I currently have with social networks.

Third, the integration with Gmail means that, much like Mozilla's Raindrop project consolidates communication, my inbox is replaced by a more generic inflow.

the massive fragmentation problem that I currently have with social networks.

I've always thought that the tendency to view social network fragmentation as some sort of problem, rather than as a rather useful feature, is a major blind spot for programmers. This Buzz debacle does nothing to discourage that hypothesis.

I blame the database training. After you've spent enough time wrestling databases you begin to dream of putting the universe into normal form. You envision a world where every piece of data might be seamlessly linked to any other with some sort of super-powered JOIN command. This is the dream that leads people to think that single-sign-on across the entire web would be a great thing. Unfortunately, this dream is also Orwell's nightmare.

I have a lot of friends. Some of them are elderly former carpenters who have never used computers. Some of them are college-aged aspiring hackers. Some know me personally. Some know me mainly as a pseudonym on HN. Some of them are middle-aged hippie folk musicians. Some of them work for major military contractors. Some of them are professors on the tenure track. Some of them barely have a high school education. Some of them are millionaires. Some of them are unemployed. Some of them are theoretical physicists, but many more of them loathed high school physics.

I have no particular desire to put all of these people in the same room at the same time. The one occasion where I was compelled to approximate that even slightly -- my wedding -- was the occasion of more than a little stress, despite every single guest's friendliness and good cheer. Because it's really very difficult to throw a party where everyone you know feels equally welcome.

I would not be surprised to discover that social networks are going to remain fragmented because most people like them that way.

No, but fragmentation with respect to single individuals is devestating. Do I send someone an IM, an e-mail, a Twitter DM, a phone call, a voicemail, an InMail, a FB message or poke, a letter, a buzz or what? If I have confidence that the target of my communications will notice my message regardless, then that is a feature to me. Moreover, if I want to follow a friend of mine, then I rarely want to setup and maintain another account, as is often required these days. Within a given social network, there's lots of room for clique-forming. There's even room for fragmentation between networks, so long as communication between networks can happen in a relatively transparent fashion.

Separation of control can't be over-emphasized. When FriendFeed was acquired by Facebook, I decided to go in and manually delete a bunch of information I'd intentionally kept off Facebook, but didn't mind having on FF, and to completely delete my (rarely used, but existent) Facebook account.

To me, it's important to have separation of control between the people that I know IRL (e.g. old high-school friends with whom I have to sometimes "dumb down" my natural style, at the risk of sounding like a haughty intellectual) and the people that I communicate with online, people with whom I'm interested mainly in sharing ideas / the interesting nerd info. The contexts are such that it's sometimes important for me to use proper logic, grammar, and syntax, and sometimes not.

Thanks! I didn't mean to be confrontational, and I appreciate you recognizing that.

Do you think Buzz could replace your FF usage, or do you think they're serving different purposes?

I didn't take it as confrontational. I've seen a lot of people mistakenly comparing Buzz to Twitter and being let down, but I think that FF is a much better comparison.

As for replacing FF, it's still pretty far from doing that. There's a lot of sites it doesn't connect to yet, and it doesn't offer tools like "rooms" and "invisible friends" that help a lot in organizing social interactions. Right now, Buzz feels like FF lite or something, and excels mainly for its tight integration with Gmail.

>> "Right now, my first six tabs are Gmail, FriendFeed, Facebook, Twitter, Voice and Wave"

Jeez. Do you manage to get any work done?

Haha ... if only there was an extension that aggregated everything, right? You'd be able to turn that six to just one. :D

This just shows how poor Google's think tank model of operation is. I mean how can all those people working on the Buzz team overlook this? With all their time spent analyzing which shade of blue they should use why not just ask people how they feel about an auto following model?

Yes, I think in this particular case although it's admirable they moved quickly to correct the major problems this is definitely revealing a weakness in Google's ability to understand humans and their "irrational" reactions to products.

I posted in an article about Google's data obsession that Google products "have no soul." (And was downmodded!) This is an example of what I meant. They are so data driven and engineering focused that I think it was inevitable we'd see them release a product that consumers freak out about.

Putting your toe in the water and using data to iterate your way out of problems works great for many of the problems we're trying to solve. But, this problem, they decided, required a "big bang approach", since invite only won't generate the network effects. So, you had to get it right on the first try. For problems like this, you need a visionary decision maker with an understanding of human beings and intuition as to how they will react to the product. Someone who puts his or her foot down without being able to necessarily justify why.

Google hates intuition, since it's not data driven and is just an opinion. Better to try it, measure it, and prove it. Most of the time, this doesn't burn them. And in fact, it'd be impossible for Google to do what Google does at the scale it does it without this philosophy being burned in deeply to the culture. But this also makes it really, really hard for them to go into social networking with a "big bang" as they tried here, since social networking is largely about people, emotions, egos, and feelings and connecting users to other users in a way that triggers the right responses and network effects. It only partially falls back on data & algorithms to make things work well. (Arguably, Facebook got the former part right first, and bolted on data-obsession later, which is why they are winning this war.) Ie, it's not that great of a situation if you have to machine-learn your way towards realizing that someone's ex-husband doesn't belong on their friend list. You've then burned the bridge and it's too late. Better to have someone with common sense step in and avoid the fire.

So, I don't think this will be the first time we see Google have to backtrack like this. It was the perfect storm for their major achillies heel to be exposed, and I'm afraid there's not much they can do to correct the root problem since it is a cultural viewpoint that would cause the other areas of their business to suffer.

Wait invite only won't generate the network effects why would you think that? Facebook and other social networks have adopted the "all our info belongs to us" method of propagation, but it's far from the only path to networks effects. Most of the major social websites have have blunders like this, and while they discourage people like me from using them, the social norms around privacy are changing.

Yeah I retracted that slightly in my post, sorry if I stealth edited it later than I should have -- I actually am not going to argue that the "big bang" was necessary for Buzz's success, only that Google seems to have made that decision.

I agree wholeheartedly

The worst part about is that it was a problem of judgement, not some technical oversight. Recently there was another article about retrieving a gmail address from a Google profile URI, now that's a technical mistake. But with Buzz you had to have whole teams of engineers thinking this was a good way to roll it out. Unbelievable.

That was an impressively quick response for a large company.

It's really not surprising though, imho, because no one likes having decisions made for them.

One of my teachers once told me what I thought and why I did things. I've always hated my teacher since that time- not because he was completely wrong, but because the principle of the action was disgusting. Google just decided we (gmail users) wanted this.

I liked those features. I liked not having to click to follow people and be automatically signed in with everything already configured. But I don't think everyone else did. I think we have different viewpoints and enough people got pissed off for them to change their policies.

People use services in very different ways, and have very different views on what is acceptable privacy-wise. Personally, I struggle with the "all-in" nature of Facebook (and Buzz, but there's no social pressure to use it yet) - I have various social circles offline and online that simply don't overlap (e.g. family, friends from school, friends from university, friends in the various towns/countries I've lived, customers/business partners, etc.), and the tone, topic and nature of communication with each group is different. I'm not talking about "secrets", I just don't think it's appropriate for them to be forcibly stirred into a pot, and people don't need to know about everything without even trying to dig it up.

I even resent the public friends list on Facebook, but at least I get to decide who lands there (unlike the initial Buzz rollout). As a result, I don't use Facebook actively and just add people as friends because it's perceived to be rude if I don't. I've noticed I'm not the only one - there seems to be a pattern: someone discovers Facebook (note that it's only recently become popular in Austria; it was practically unknown outside the web scene in 2008), get excited and add all their friends, post pictures, etc. and after a while remove a lot of personal details, delete photos and only participate actively by replying to others. Maybe this has already played out a few years earlier in the US.

Oddly enough, I really enjoy Twitter. I guess it's because it's very up-front about everything being public, so the temptation to post something you might regret later is reduced.

It's one thing to decide that your users want something, but that doesn't mean it happens. You have to decide how you are going to fix it and then write a fix and then test the fix. Usually big companies require a group of people to agree on what the fix will be and how to do the fix. It often involves paper work and approval on different levels. It isn't usually just something that one of the programmers goes in and fixes (although that model seems to work often for startups).

Not for the level of freak-out that people had about it. You can light a fire under a company's butt if they think that the perception of their new product threatens their business, as a reputation for recklessness with user's data and privacy threatens Google.

The title is wrong. Buzz itself is not opt-in, auto-follow is, but not yet, and not retroactively.

The blog post also glosses over something critical:

This created a great deal of concern and led people to think that Buzz had automatically displayed the people they were following to the world before they created a profile.

But what about people who already had a Google profile, who did have their contacts automatically displayed to the world? Google still hasn't acknowledged this or apologized for it, and they continue to act like everyone who is complaining is getting upset over perceived rather than actual privacy concerns.

"But what about people who already had a Google profile, who did have their contacts automatically displayed to the world?"

One of my accounts had a google profile, and nothing was displayed automatically; it still required the exact same steps as the accounts without a google profile, including the "How do you want to appear to others?" form with the checkbox to display followers/following lists.

I've personally seen it happen with 3 separate accounts, including my own. I don't know what we did different than you - we logged in to gmail, clicked 'no thanks' on the Buzz pop-up, noticed it was turned on anyways, checked our already existing public profiles, and saw a publicly available following and followers list.

I'm impressed they've been able to adapt it this much so quickly. Most large software companies, at best, take weeks to make the changes that Google made. It is also good to see them listening to the feedback, you don't always see that either.

Well...it is the least they could do after their blatant disregard for the privacy of its users!

Amazing how dramatically they've had to rework Buzz in just a few days. Google will think twice about rolling anything out to millions of people at once from now on.

I actually think their approach was pretty good. You can't usually guess what users want. The guys at Google probably thought, "yah people will really like it to automatically link without any work". It's tough to gauge what the users want until you put it in front of them. It may hurt sometimes at the beginning, but if it leads to a better product in the long run then I would say: keep it up.

No. It is unethical to risk user privacy and data just to benefit the company.

"We quickly realized that we didn't get everything quite right. We're very sorry for the concern we've caused and have been working hard ever since to improve things based on your feedback"

Not the best possible apology (it's closer to "sorry you were concerned" than "sorry for our mistakes"), but not shabby as these things go.

From reading the article though, it was people's concern of things, not Google's disregard for them. The changes basically amount to making things easier to find. Even the Auto Following thing was controllable according to the article.

So, as far as they are concerned, they didn't violate privacy as much as make it easy for people to by-pass the privacy protections in place.

Both of Google's posts on the issue talk about notification and controls for users whose profile is created when they first use Buzz. But from my experiments, if you were in the minority of users who had already set up a Google profile (which I did in 2007 when they were first announced[1]) then Google started broadcasting information about who you were following before you even knew that you were following them.

Adding new privacy controls is not nearly sufficient to protect privacy; you need to notify users about them and let them choose before applying the settings to their account. At least during the first launch (I haven't re-tested with the latest changes), for my test account that already had a profile, this was the experience:

(1) As soon as I logged into GMail and saw the Buzz welcome screen and clicked "no thanks," my account began "following" people chosen by Google. (2) This information was broadcast to a group of people chosen by Google. (3) I was not informed that either of these things had happened. (4) New privacy settings were added to my "Edit Profile" screen, but (5) the least-private setting was activated automatically and I was not told yet that the new settings existed until after information about my email contacts was broadcast to other users.

I don't think anyone disagrees that there were privacy controls. But it was really a "beware of leopard" situation, at least for accounts in this particular state. The only option for these users to preserve their privacy was for them to (1) magically know that before logging into GMail today, they must edit their profile to be non-public, and (2) magically know that new privacy settings had been activated in their account so they could change them before trying Buzz.

Now, I actually like Buzz and am still using it[2]. I think it's a much-needed challenge to Facebook's closed platform. I can see how the Buzz team didn't forsee just how bad this would be for a very small minority of users. But to pretend that all users were given a chance to protect their privacy and just didn't take it is really not true. Some users' privacy was violated the instant they logged in, and giving them tools to help them clean up later (and not telling them about the tools!) does not even remotely help.

[1]: http://techcrunch.com/2007/12/15/google-starts-rolling-out-c...

[2]: http://google.com/profiles/mbrubeck

You make some good points. I won't repeat, or argue, as I agree with most of what you said. My main contention is that there was a lot of reactionaries proclaiming things that simply weren't true and were making grand assumptions. It made for an environment that was difficult to question. At some point, I was told that readers of Hacker News essentially could not be wrong, and that I was wrong for asking questions or disagreeing.

Thank you for providing more information. =)

Most of the damage has already been done. They've lost on the initial buzz that Buzz would have otherwise created. While I appreciate Google's swift response, they made a fatal error during the rollout which will cost them heavily in the long run.

You mean in the short run. People will kind of forget this after the techblogs and news sites see a more interesting story. Like now, I'm sure people are still suffering very hard in Haiti but media coverage dropped a lot and i don't see many people talking about it.

It has done long run damage to the Buzz product though. First thing my girlfriend asked me when logging into her gmail account this morning was "how do I turn off this Buzz thing". She didn't know what it was, only that there was a lot of 'buzz' on the web about how you should turn it off. I have a hard time imagining a scenario where all these people who've turned it off will somehow change their mind and turn it back on.

Perhaps, but I think that even in the long run, people will have lost confidence in Google as a brand, even if they don't remember the particular reason why. Ironically, this damage can propagate along the same social graphs that Google tried to exploit. Users who have lost confidence will be less likely to recommend Google in the future, meaning less people at the second order recommend Google, etc. Even after the story has forgotten, one can well imagine the social graph itself remembering.

This title is written in past tense but from reading it I'm told what Google is "going to" be doing.

I was going to say "too little, too late" but it would be more accurate to say "just enough, too late."

Well, quick response because they realized they screwed it up...

Sorry if this is off topic, but: the speed that Google changed Buzz is the reason why I love the web as a deployment platform. Except for local software development tools, Latex, and video editing software, I want everything I do to be web based.

This is nice to see, but in my mind, the damage has already been done. I felt really violated by Buzz's rape of my Gmail account.

It will take a long time to heal those wounds

I understand your concerns, but the use of the word rape is at best rather loaded. One could make the argument that Google abused your data, but to invoke a metaphor to extreme sexual violence dilutes your own point as well as disrespects the very real victims of rape.

The word "rape" has multiple meanings. At least one of which applies and is not necessarily connected with anything of a sexual nature. http://en.wiktionary.org/wiki/rape

Yes, it does. The trouble is that the connotations of the word are still very highly sexualized, even when using an explicitly non-sexual definition.

The legal definition of rape hangs on the detail of lack of consent, not on how violent it was. It's kind of a myth that rape is "extreme sexual violence". In the majority of cases, it isn't all that physically violent, but it is still a violation of the person due to lack of consent. In that sense, "rape" makes a good metaphor for what Buzz did here.

Buzz is now how it should have been all along, but to be honest, like parent, its unliekly I'll be re-enabling Buzz again any time soon.

Reply on news.ycombinator.com