From Issuepedia
Jump to navigation Jump to search

This discussion of Google corporate policy is excerpted from comments on this thread.

The discussion basically began when I said:

...Google has never given any genuine indication of being the slightest bit interested in my suggestions or in the suggestions of anyone I know. If they were, there would at least be a bug-tracker of some kind so we'd know whether our suggestions were even being considered or not.

To which Regalo replied:

There is a public bug tracker for Chrome, but for most non-open-source projects they keep that stuff hidden so nobody has to be afraid of leaking confidential information when talking about bugs.

My view from the inside is that people at Google care a great deal about making the best products they can, but practical realities get in the way a lot. Sometimes Google does stupid things because of an executive decision (e.g. real names), but usually it's just a lack of resources. Google is a big company, but it has a huge number of products, and only so many people can work on the same product at once, and there are often fierce debates about the best way to fix any given problem.

Google typically doesn't respond to feedback because there's no point; the problem is usually already known, and your feedback is basically a vote fixing that problem before working on something else. Our policy is not to discuss work in progress or commit to releasing things at specific times, so there's not much to say until a bug actually gets fixed.

At this point, the conversation became multi-threaded, so here is my attempt to represent that (added comments -- i.e. not from the G+ discussion -- are in italics):

  • RC: ...but for most non-open-source projects they keep that stuff hidden so nobody has to be afraid of leaking confidential information when talking about bugs.
    • WH: In other words, the profit motive keeps the developers from doing the right thing.
      • RC: Well, most Googlers do like to eat.
        • WH:
          • There is a difference between corporate profits and employee salaries. Even nonprofits pay their employees.
          • You can't be a force for good and place profit over doing the right thing.
          • At around $60 billion, Google's revenue is larger than the economies of many countries. When an organization gets to be that size, other considerations should take priority over short-term profit.
          • And finally, it seems to me that long-term business health is harmed by opacity -- unless Google has (as many of us suspect) truly given up not being evil [and decided to focus on being profitable].
  • RC: Google typically doesn't respond to feedback because there's no point; the problem is usually already known.
    • WH: Again, why isn't there at least a searchable listing of known problems, with status indicators? That wouldn't risk divulging proprietary information, since all text would be user-submitted.
      • RC:
        • User feedback is confidential.
          • WH: That was never a decision made by the users. In fact, if I remember correctly, users are given the option to mark out any areas of the screen that they wish to keep confidential. Why is the remainder not considered public? Why is there no option to "make my comments public"? There could even be an option to make them anonymous.
        • If you want to share your feedback through other channels, they're already available.
    • WH:': The problem may be known within Google, but those who submitted it have no way of knowing that -- much less whether Google intends to do anything about it. Google may even consider our "bug" to be a "feature", or that fixing it is not consistent with their business goals. They may not see the point in a feature or fix that many people would like, and we have no way of knowing that we have not explained it clearly enough. They may be working on it -- but from outside, it looks like nothing is happening... except the usual "improvements" that often make things worse.
  • RC: ...and your feedback is basically a vote fixing that problem before working on something else.
    • WH: The list could show how many "votes" there are for a given fix, so I'd know whether I was barking up the wrong tree or not.
      • RC: In theory, but when I've used systems that support this kind of feature, I haven't found it terribly helpful to me as a user. It's also hard to do in an automated way, because bugs and feedback reports don't match up 1-1. One bug can cause several issues, or several bug could cause the same symptom.
        • WH: These kinds of systems are in common use for both proprietary and open-source development. Bugzilla, Redmine... The last place I worked made extensive use of JIRA for tracking bugs in their extremely proprietary and closed-source system. Are you telling me that Google has no internal bug-tracking system?
  • RC: Our policy is not to discuss work in progress or commit to releasing things at specific times...
    • WH: Why?
      I'm not really expecting an answer to this. I have a close relative who was on the G+ privacy team (something like that) around the public launch time, and we spent about 2 hours going back and forth about Google's policy of opaqueness.
      • RC: It's not something I or anyone else you might talk to is in a position to change. Our policy, handed down from the top, is to favor internal transparency at the expense of external transparency. This puts people like me in a position of trying to defend Google based on facts we know but can't share.
        • WH:: ...which is not a position I envy -- but I have to judge based on the facts available to me. I'm not willing to trust that decisions made in secret are for my own good unless there is some kind of post-secrecy accountability process, nor should you (it seems to me) be willing to entirely trust your own perceptions of company decisions, when the company in question can easily overwhelm your objectivity by the way it frames and narrates the context of those decisions.
      My position was (and is) that opacity just makes each leak that much more problematic -- whereas if the overall process were transparent, nobody would care. This attitude seems to be so deeply ingrained in the corporate culture that Goog insiders can't even see how destructive it is; it's the fishtank effect, where it's difficult to conceive of the idea of "water" when you're surrounded by it. Anything else seems alien.
      • RC: Perhaps, but it's how all of our competitors operate.
        • WH: To the best of my knowledge, all of Google's competitors are unabashedly evil.
        • WH: And... seriously, "all the guys are doin' it" is hardly a defense.
      This is what will ultimately kill Google: the rest of the world will move on to open-source, distributed systems, while Google will still be trying to be Microsoft Mark 2.
      • RC: We talk a lot about what will kill Google, and turning into Microsoft is high on our list of worries.
        • WH: Good. I'm glad to hear that. You are no doubt aware that similar market forces are going to shape companies in similar ways, and it would be all too easy to travel down the same path that Microsoft took: slowly morphing from a tool-finder (Microsoft didn't write DOS) into an eater-of-ideas. If Google is to avoid the same fate, it is going to have to make a conscious decision, at the highest level, to choose a different path. Is the highest level even interested in this? Is there any kind of internal campaign to develop such a plan and ensure that the execs engage with it on its own merits?
      Either that, or Google will take over everything so completely that there will be no escape from the centralized control it increasingly enforces on its users.
      • RC: Not gonna happen. Google is not immune to anti-trust litigation, and even the appearance of a monopoly is a huge threat to our business. We recognize that maintaining a healthy ecosystem involves working with competitors and not trying to sabotage them. There will always be alternatives to every noteworthy Google product, and Google itself has every incentive to avoid totally dominating any niche. You may be temped to bring up something like Reader as a counterexample, but consider for a moment the reasons why this product may have been killed.?
        • WH: Then why is Google increasingly forcing integration of all its services? The YouTube-G+ integration is just one example. This could have been done in a way that kept the services separate (via an open API) and allowed users to control how (or if) their accounts were connected. The tight integration of Android devices with Google services is another example. What if I want to sync my photos and videos to my home server instead of to Google's servers? I have to install a separate app for that (which I wouldn't even know about if I weren't heavily into Linux); I can't just reconfigure the existing sync app.
        • WH: Google has a gross revenue of around $60 billion. This is approximately four times the (unofficial) budget of the CIA. This is larger than the GDP of many countries. It is already far too big for a private company. (And yes, the CIA is too big as well.)
      • RC: A variety of reasons:
        • Sometimes we don't want to broadcast our plans to competitors.
          • WH: This may seem like a naive question, but: why not? Is it really that important to be first out of the gate?
        • Other times we just don't want to open ourselves up to accusations of vaporware.
          Certain other companies like to announce their products well in advance in order to build up hype and perhaps dissuade people from buying competing products that are already on the market. With rare exceptions, Google does not play this game. We try to release products and updates when they are ready and not when our marketing teams announced they would be ready. Stuff still gets released too soon, but it's mostly because people who've been working on a product forever are sometimes too eager to ship.
          • WH: If open discussion of products under development -- or even under consideration -- were routine rather than "hot leaks", nobody could legitimately make that accusation.
        • Sometimes there are also business and legal reasons why certain information can't be shared. Patents and licensing agreements make everything much more complicated. Sensitive negotiations with outside parties are usually hush-hush even inside Google, because the outside parties rarely want to share what is going on until a deal is finalized, and they want to be able to control their own messaging.
          • WH: I can see how legitimate and well-intentioned business decisions could have led down this path, but it seems to me that the pervasive secrecy to which those decisions have led is sufficient cause for re-evaluation of those decisions.