Improving Stakeholder Communication in Firefox

As I’ve been watching this rapid release thing continue to unfold, I’ve been thinking about what it is that truly went wrong in the messaging. The fact that at this point, Mozilla executives are having to continue to try to sell this to people means that something went wrong in how they executed the change. As I thought about it more and more, I realized what a big part of the problem was – stakeholders.

If you read any books about successfully making change in your organization, they bring up the important of getting the stakeholders on board with the change early on in the process. As I look through the history of Firefox, I see over and over again where Mozilla made major changes and failed to communicate with the stakeholders before making changes. Here are just a few examples:

Change: Personas
Stakeholders: Theme developers

When Firefox introduced the concept of Personas, they failed to account for incompatibility with themes. This continues to be a problem to this day. As a result, they completely alienated most theme developers.

Change: Rapid Release
Stakeholders: Enterprises, Add-on Developers, Localizers, Bug Triagers

When Mozilla introduced rapid release, they failed to account for additional work for localizers, bug triagers and add-on developers. They failed to account for enterprises that have dependencies on longer life cycles.

Change: Third Party Add-on Notification
Stakeholders: Add-on Developers

When Mozilla announced third party add-on notification, they failed to account for the shear amount of third party add-ons that are out there. According to Mozilla, a full 75% of installed add-ons are NOT from AMO. When this change goes in, it’s going to seriously impact the add-in community.

Change: Removal of Firefox Community Editions
Stakeholders: People Repackaging Firefox

When Mozilla removed Firefox Community Editions, they did it silently. Simply turning the page into a 404. They didn’t even realize what they’d done until someone told them.

Change: Removal of Status Bar
Stakeholders: Add-on Developer, Users

When Mozilla removed the status bar, they failed to account for the fact that it would break so many add-ons. Luckily sanity came in and they got something to fix it before Firefox 4 shipped. But users continue to complain about the loss of the status bar.

Change: Removal of Version Numbers
Stakeholders: Everyone

I realize this change was reverted, but this is another classic example of just throwing something out there without really gauging what the impact would be to the community.

 

Recently there was a blog post on planet that talked about remembering that Mozilla was not a company supporting a community, but a community supporting a company (or something like that). That might be true with the Mozilla organization in general, but when it comes to Firefox, they behave more like a company that makes decisions without really regarding how they will affect the rest of the community. Or maybe they do think about it but just do a very poor job communicating it.

The argument is generally “we discussed that in the newsgroups,” but I don’t buy it.

We live in a modern age where people communicate in blogs rather than newsgroups, and this is an area where Mozilla falls way behind. The main Mozilla blog doesn’t even allow comments, and when the add-ons blog is used to communicate changes, they are always posted after hours west coast time. So there are no responses from Mozilla to questions until usually a full 12 hours after the posting. Announcements like this should happen early in the day on the west coast so there can be an active discussion. They also do a very poor job moderating, with comments taking as long as 24 hours to show up.

In the end, despite all the information that comes out of Mozilla, they do a very poor job actually communicating. Communicating is not just about producing words or announcing things, it’s about having an interactive dialog. And lately (especially as an add-on developer), it feels like most of the time it’s just Mozilla saying “this is the way it’s going to be.”

That’s not how you get people to buy-in to your ideas.

Please note: I reserve the right to delete comments that are offensive or off-topic.

Leave a Reply

Your email address will not be published. Required fields are marked *

29 thoughts on “Improving Stakeholder Communication in Firefox

  1. Agreed. At first I thought it was just AMO but it’s really bad across the board – and it’s frustrating for everyone afterwards because the folks pushing through the decisions are stunned to find that even keen stakeholders are missing the boat. They get impatient, feeling like they’re retreading ‘old ground’ for the sake of individual parties; when really the issue is that those individual parties, those stakeholders, should have been (and wanted to be) involved from the beginning.

    In spite of my interest in so much Mozilla I refuse to use the newsgroups. Not because of some crazy ideological bent but because there’s so much useless information. I could not possibly consume relevant information in any fashion from the Mozilla newsgroups, I’d burn out before the week is up. They’re full of junk and I’m not paid to spend my entire day looking through Mozilla groups for things that might be relevant to me.

    Most of the time stakeholders don’t even know these discussions are going on until the feature hits the patch list or The Reg run a complaint article. Planet Mozilla is too broad and doesn’t try to represent everyone, it’s not a suitable alternative.

    The solution, to me, feels like some kind of portal where you subscribe to receive information about your chosen subjects. Not in the form of automated discussion digests – but human-powered summaries of the who, what, why, when and where of everything Mozilla for that subject over the current point in time which direct you to places of action:

    “We’re looking at where to go with [a]. Want to help? E-Mail [x] and join the discussion over at thread [y], we’re also going to be discussing it at conference [z] if you’re free!”

    Anyway – musing out loud again…

  2. Another issue is what is just an idea, and what is set in stone. For example, removing the Firefox version numbers. Asa made it sound as if the whole change was set in stone and was happening anyway. Then Faaborg came along and said it was just an idea. If that had been said before the discussion, it might have been a bit less heated.

  3. > They failed to account for the shear amount of third party
    > add-ons that are out there.

    You have cause and effect backwards. The change is being made _because_ there are so many third-party add-ons out there and because the vast majority are somewhere between malicious and broken. This change is being made to protect users from the very people you list as “stakeholders” for this item!

    • > This change is being made to protect users from the very people you list as “stakeholders” for this item!

      I completely disagree. I think Mozilla is throwing out the baby with the bathwater with this change.

      Mozilla is misusing the term “third-party” in this context. There are add-ons installed outside of the profile. There are add-ons installed via mechanisms other than XPIs. What Mozilla is doing is demonizing add-ons that are not installed via the normal way (which in a lot of cases is a very legitimate way to install them).

      This change is not just about third-party add-ons either. This change encourages users to disable add-ons without giving them any information except “add-ons can make Firefox slower”

      This change does an incredible disservice to add-on developers.

      If Mozilla finds add-ons that specifically impact Firefox in a negative way, it should blocklist them and work with the add-on developer to get them fixed.

      Making blanket statements about all add-ons and all “third-party” add-ons is just wrong.

  4. I have raised the issue in IRC a few times… that Mozilla does not have a “news” page… whatever announcements are made are just sorta thrown in the blogs and planet mozilla…. From what I gathered, they got miffed at having stuff on a news page because no matter what was posted on there, it was always misinterpreted anyway.

    I guess the current thinking… is to do whatever they feel like, and if there is backlash, then pretend to be concerned and back out if the news hits slashdot.

    Most everything is thoughts thrown out there, and then there is some thumb twiddling to wait and see if it was a serious post, or just some “post to start a discussion”

  5. Amen, and double amen to Tyler’s post. People understand that not all decisions can be made with an extended public input period, but too often it seems like all ideas are set in stone before anyone outside of inner circles hears of them.

    On the “we discussed that in the newsgroups” issue, I just aired my opinion in response to such a claim by Asa over at http://www.gavinsharp.com/blog/2011/08/29/thousands-of-bugs/ ; perhaps that may develop into an interesting discussion.

  6. I hope *internal* communication is better than *external*… FWIW, I sent a message entitled “Important questions regarding my company and Mozilla” to Gary Kovacs, Mozilla’s CEO on the 13th of july. Pinged him again with a resent on the 20th. Then again a few days later through twitter, he pinged back through twitter too but I got no answer to my email. Pinged him again on the 1st of august, Gary then requested yet another resend… Resent… Still waiting for an answer, 48 days after my original message.

    This is not the first time I send a private message to Mozilla’s CEO between 2003 and now, like I sent a few private messages to Netscape’s VP before 2003. I never dare doing that without an important reason. In the past, I always got an answer, from Netscape’s VP or Mozilla’s CEO. Not necessarily a positive one (from my own perspective) but I _always_ got an answer and that’s what does matter. So this is the first time I receive no answer _at all_. Hum…

  7. > We live in a modern age where people communicate in blogs rather than newsgroups, and this is an area where Mozilla falls way behind.

    I think that all open source projects I can think of have most of their serious debate on newsgroups, not blogs. For example, the Linux kernel does that, as does Ubuntu, etc. Mozilla is no different there.

    Newsgroups are much more convenient than blogs in many ways – there is a single ‘interface’, as opposed to blogs where each one needs a separate user account, separate syntax for editing content, etc. A newsgroup is much more centralized, you can find the entire debate instead of looking at who knows how many blogs are mentioning a certain topic.

    Blogs might reach a wider audience, and for that matter so does twitter. These things have their uses, for sure. I agree with you that it would be a good thing for Mozilla to focus more on responding to feedback in those locations. However, serious debate should be like in other open source projects, which is to say mainly in newsgroups. They are not as cool but they are effective and part of open source culture.

    Just my personal opinion of course.

    • As a friend of mine told me:
      “Newsgroups get caught up in ‘Person A starts thread, B, C, and D respond. D doesn’t get a reply because he either agrees or makes a dumb joke, B and C get replies from A. B replies to what A said to both B and C, and E and F reply to B. G comes and shouts that he doesn’t agree with anything and makes useless threats. H reprimands G. A replies to B. I replies to E. F replies to I. A replies to both I and F. E comes back and asks the same question and reiterates a question B asked five emails ago that A didn’t answer. J comes and replies on A’s behalf since it is 5:00a on the East Coast. A clarifies J at 10:00a. C replies that he can’t follow the thread. H summarizes. G complains again. A reprimands. C disagrees with A. B asks H what his opinion is. H asks a question of F. F doesn’t reply. A starts arguing in a subthread with F over another problem, and that’s where I stop reading the thread.'”

  8. Thi page needs a +1 button to go along with all the other social network buttons … mainly so we can see how many digits Google allow in the counter before they start using exponent notation 🙂

  9. I think you’re conflating failure to communicate with “outcome wasn’t what I wanted” on most of these issues. While I agree there’s a lot we could do to communicate better, I disagree with most of the examples here. To address a few that I’ve worked on:

    > When Mozilla announced third party add-on notification, they failed to account for the shear amount of third party add-ons that are out there

    We’re well aware that there are hundreds of millions of add-on installations not from AMO (as you note, we’re the ones who said that) and we still think it’s important enough to move forward with. Nearly everyone who uses Firefox on Windows has to deal with third-party-installed add-ons forcing themselves in Firefox, and giving users control over that is well overdue. I realize there will be tons of add-ons disabled as a result of this, and that’s exactly what we want: the user to be able to decide what add-ons are running and affecting their security, privacy, and performance online.

    This wasn’t a case of us failing to account for it — we accounted for it and realized how big of a problem it is.

    > When Firefox introduced the concept of Personas, they failed to account for incompatibility with themes.

    Personas have brought customization to tens of millions of users who didn’t know about it before. There are over 300,000 Persona designs submitted to us tens of thousands of designers, some of whom have told us they never would have imagined being able to create something and see other people around the world using it.

    Yes, users must choose between having a Persona and a theme. Mozilla decided that it was more important to bring this functionality to users than to wait for the new feature of theme compatibility with Personas. This wasn’t a failure to account for it, it was a logical decision that we shouldn’t hold back an awesome feature because a very small number of users will want to have both a theme and a Persona. Before the launch of Personas, around 7% of Firefox users were using a non-default theme. Today, more than 20% of users are using a Persona.

    > When Mozilla introduced rapid release, they failed to account for additional work for … add-on developers.

    Add-on compatibility has always been one of (I’d say THE) biggest hurdles to shipping Firefox faster. Of course we knew it would be extra work, which is why we worked on a plan and discussed it with the community for months in advance of rapid releases. The plan was for AMO-hosted compatibility, and it’s gone really well. 97% of add-ons compatible with Firefox 5 were compatible with Firefox 6 on launch day.

    What we failed to account for was the huge number of add-ons not hosted on AMO that would make people think their add-ons aren’t compatible, when it’s really just 100 million incompatible Java Console installs that shouldn’t exist in the first place.

    More than 90% of several hundred add-on developers surveyed last week knew about rapid releases in advance of Firefox 5, and most of them have not had to make any code changes for their add-ons to stay compatible.

    > when the add-ons blog is used to communicate changes, they are always posted after hours west coast time.
    I’m fairly sure this isn’t true, but more importantly, Mozilla is global and we aren’t going to establish an official policy of posting blog posts that are convenient for a certain timezone, and if we did, it probably wouldn’t be for the US as most add-on development takes place in Europe.

    > They also do a very poor job moderating, with comments taking as long as 24 hours to show up.
    I personally visit the Add-ons Blog moderation page multiple times a day during the week and at least once on weekends, and others on the team do as well. Users who have previously posted are automatically approved unless they have more than 2 links in a comment. I don’t get this comment at all.

    Making product decisions about software used by hundreds of millions of very diverse people isn’t easy. Someone will always be upset. We do the best we can to minimize that. It’s easy to constantly be negative and write blog posts throwing stones and making assumptions. People that work at Mozilla are pretty smart and it’d be great to have the benefit of the doubt that we’re not so incompetent and reckless that we’re failing to account for who is affected by what we do.

    • Sorry, I confused two topics with the third-party add-on thing.

      My bad.

      My earlier comment to bz contains my opinion of the the new third-party add-on dialog.

      As far as Personas go, to date, Mozilla has still failed to deliver on the promise of Personas/theme coexistence. That’s the biggest thing theme owners complain about. (And as I side note, I’ll point out that Mozilla is in the process of alienating Personas developers my reducing Firefox chrome to a point where you won’t be able to see a Persona at all 🙂 )

      As far as add-ons and rapid release, you actually confirmed exactly what I said, which is that you underestimated how many add-ons are not hosted on AMO.

      As far as posting to the AMO blog goes, that is absolutely true. Feel free to turn on timestamps and you’ll see.

      You announced the third-party add-ons changes late at night, you announced AMO slugs late at night, and there were many others. I understand we live on more than the west coast, but when you announce a major change, you should probably do it at a time where you can actually be around to monitor the response.

      And I’ve watched my own comments sit in moderation for a day.

      • > That’s the biggest thing theme owners complain about.

        In my experience the biggest thing theme owners complain about is that Personas get more promotion on Mozilla websites than themes do, not incompatibility with them.

        > As far as add-ons and rapid release, you actually confirmed exactly what I said, which is that you underestimated how many add-ons are not hosted on AMO.

        That’s not the context you said it in. You said we failed to account for third-party-installed add-ons in regard to the third-party notification process, which is not true. We had the data when we made the decision to go ahead with that.

        I said we failed to account for it in rapid releases, which was because we didn’t have any data on it until after Firefox 4 was released. Once Firefox 4 was released and we had the data showing what the top add-ons are whether hosted on AMO or not, the rapid releases plan was already under way.

        > You announced the third-party add-ons changes late at night, you announced AMO slugs late at night, and there were many others.

        http://blog.mozilla.com/addons/2011/08/11/strengthening-user-control-of-add-ons/ was published on Aug 11, 2011 @ 11:37am

        Slugs and other AMO releases are usually posted in the evening because AMO releases used to be at 4pm, which means we weren’t ready to announce it until 6pm, and we’d rather post it that night than wait until morning when people might be confused. Releases are now at 2pm and are much faster to deploy, so there’s a better chance it will be announced in the afternoon.

        • I’ll stand humbly corrected.

          I think I was confusing that date/time with another time.

          Thanks for the AMO information, that’s helpful.

  10. Daniel: with all due to respect, but gauging the project internal and external communication health and organization internal and external communication health by your personal emails to a CEO of said organization is rather prone to really bad conclusions.

    I generally agree with you, Mike, that in cases he listed Mozilla (both as a project and as an organization) could have done better job building communication and feedback loop early on, but:

    1) You mixed multiple elements into one bucket “poor communication”

    Some of that may be Mozilla not being able to communicate well, some may be that Mozilla failed to build the right communication channels in time, some may be that people who worked on the project had no experience in fitting their project into the Mozilla ecosystem, some may be intentional decision-making, some may be that there are bugs that are overrepresented in your sample, but not that meaningful in the grand scale of things, some may be that you tried to impact the project using the methods that worked in 2003 but does not work anymore, some of it may be that Mozilla communicated so many things that catching this one item that is crucial to you or me may be harder than it used to be, so our perceived noise/signal ratio is higher, but each of us is looking for different areas etc.

    2) Change in the project requires us to rethink how we enable contribution and how we contribute

    I feel that one of the outcomes of the growth of Mozilla project and diversification of its areas of interest is that we need to relearn how we, contributors, operate inside the project.
    Old ways may become less efficient and less effective and what we perceive “oh, mozilla is not open to my feedback now” may be just because it now takes more focus to learn how to give feedback and get involved in a certain area of the project and it becomes impossible to cover “everything” like we used to.

    It doesn’t mean that there’s no issue of how open certain projects inside Mozilla are, and I believe there’s a lot work ahead of us to regain the level of openness that we had in the early days in this new world, but I’m positive that Mozilla is interested in doing it and investing in it.
    But that doesn’t meant that in this new reality contributing to the project will mean the same and look similar to how it used to look like. We all need to also invest in inventing new patterns in which we contribute to the project in a meaningful way.

    tl;dr: communication is a two way process, and we all need to invent how we can have it functional in Mozilla project 2011, not just the project->contributors, but also contributors->project. Finding your ways inside the project is certainly different than it used to be, old patterns are less effective, and I believe it’s misinterpreted by many old timers as “The grass was greener when I was young”.

    • On March 18, 2011 you asked for feedback from localizers.

      It had already been decided to go to rapid release at this point.

      The point of getting stakeholder buy-in is to have discussions BEFORE decisions are made.

      So that they can impact the discussion.

      This simply proves my point. That stakeholders weren’t involved in the discussion early enough.

  11. “What exactly is open for discussion? Mozilla itself is, but is Firefox?” is what I have been asking/wondering for a while.
    http://blog.lizardwrangler.com/2011/08/22/community-wide-discussion-of-mozilla-in-the-new-era

    Mike stated something similar above, but I’m asking.
    Having a very clear answer would help sort out a lot of things for people.

    A lot of people are saying that things are and have been presented as final and that there hasn’t been the opportunity to provide input during the planning/shaping process. It is after final decisions are made that feedback is accepted. And it’s just that, feedback, and that leaves people feeling like outsiders.
    Whether this is all true or not, and there are actual opportunities to be involved in planning, the perception again, is that what is announced is final. Even if the feature, policy, etc won’t actually be implemented for quite some time.

    People are free to go to the wiki and see road maps, plans, etc, but they’ve already been decided on.

    I know that Mozilla is in a tough position having to be open and hear 1,000 different things from 1,000 different people, but isn’t that just they way it is with an open organization? Shouldn’t the overall consensus from contributors, volunteers, end users, etc play a big role?

    A few other comments.
    Personas
    Theme developers did feel that they weren’t being appreciated or respected at all and as a result, we lost veteran themers, and long time contributors.
    AMO
    I have to disagree with you somewhat Mike.
    Jorge Villalobos is pretty good and consistent with replying and addressing concerns.
    The timing issue, I don’t have a problem with. I live in the U.S., but the World is much larger than just us here.

    Third part add-ons
    I’m happy about the decision with that.
    The option to enable third party add-ons is there.
    I can’t stand it when an app totally unrelated to Firefox tries to, or somehow successfully installs an add-on, especially when it cannot be uninstalled.

    Rapid release
    Well, I’m just not a fan at all, but I support Mozilla’s decision. I don’t like it, and I hope that they revert, or tweak it, but right now, it is the way it is.

    To end my babbling, having the answer to this would be most helpful.
    “What exactly is open for discussion? Mozilla itself is, but is Firefox?”

  12. About Personas and themes (something mentioned in several of the above comments):
    I’ve seen a lot of nice Personas (many tens, maybe hundreds), I’ve even downloaded them. But I’m not using them. Why? Because I’m using a theme which is not the default theme. I can see (but for only a few seconds at a time) how nice a given Persona would look on this time (by hovering the mouse over the Persona’s thumbnail).
    But usually I don’t complain. Not because I have nothing to complain about: if it were possible to wear a Persona with my favourite theme, I would. But because I’ve been told time and again that “whining in a bug comment leads nowhere; if anything, it makes the bug less likely to be fixed”.
    So I’ve added my Bugzilla email address of record to the CC of bug 520124, I’ve even voted for it, and that’s that.

    • Thanks, Ken, it works! 🙂 And Personas can still be previewed at AMO or getpersonas

      But, too many restarts! (three to start wearing a Persona over my current theme, and no way to change Personas on the fly).

      There used to be a pref (which never worked satisfactorily) to change themes without restarting, I looked for it to see if it would make it easier to change Personas, but couldn’t find it back.

      I’ve found the following to be useful in userChrome.css (searchInput is in the mailer, the other two in the browser, tabs in both):

      /*
      * Reveal more of your Personas by making chrome less opaque
      */
      #urlbar, #searchInput, .searchbar-textbox
      { opacity: 0.5 !important
      }
      #urlbar[focused=true], #searchInput[focused=true],
      .searchbar-textbox[focused=true]
      { opacity: 0.8 !important
      }
      .tabbrowser-tab
      { opacity: 0.667 !important
      }

  13. I’d been searching for some answer as to what has gone wrong with Firefox this year. I wholly agree that the core cause of many issues with Mozilla is communication. But at this point, so many people are loosing faith in the Firefox project, I suspect we will see growth numbers diminish over the next six months. I used to use Firefox for everything except load Oracle apps (which is Oracle failure, not Firefox), now i find myself giving in to the IE temptation due to a need to get work done, and along greatly missing functionality I used to rely on.

    A communications release process with subscription services seems like a good idea, but this is not likely to cause a paradigm change where it’s need, and that is at the top.

    If I were a conspiracist, I’d almost say Mozilla has been infiltrated by malpersons from another organization with intent to disrupt the project to it’s demise. It’s just so hard to believe that these highly intelligent people would be so daft as to not understand the importance of stability and smooth transition when evolving a product like this.

  14. You can stick with your outdated browsers and outdated development methodolgies all you want, I’ll enjoy my 30% faster performance in Firefox 9.0 with Type Inference.

    http://www.extremetech.com/computing/94532-firefox-9-javascript-performance-improved-by-20-30-with-type-inference

    http://www.conceivablytech.com/9188/products/firefox-9-gets-30-boost-in-javascript-performance

    Enterprise staff could be so much more productive and get more work done with these kinds of speeds, to dinosaurs called enterprise companies, your loss, stick to the slow lane, the Internet has moved to warp speed.

  15. Mike, you’re right that communications have been a big failing – and communication needs to be a 2-way process.

    However the problem behind that is that those at the top seem lack the appreciation for the need for good communications. While I now little of their individual backgrounds, the outside appearance is that FF is too dominated by those whose main focus is on developing the technology. I suspect that they need to widen their management team, and also learn that it is not always the best technology that wins out, but those who can carry consumers with them and / or create an industry standard. Case studies: Betamax, Microsoft, Apple.

  16. A couple of thoughts for your consideration.

    Regarding Justin Scott post:

    Your response reminds me the way Google Chrome support often reacts to questions and comments on some of their user forums. They have a tendency to be a ‘bit’ too defensive; which causes frustration and disgust for a lot of well meaning users. I think most of these users consider their questions and comments to be sincere, helpful, and well thought out.

    Regarding Mike Kaply post:

    You wrote:
    > I’ll stand humbly corrected.
    > I think I was confusing that date/time with another time.
    > Thanks for the AMO information, that’s helpful.

    A true gentleman: “You know when not to continue a debate with a politician!

    Regarding Axel Hecht post:

    You wrote:
    > If you don’t like the technical means in how
    > communication happens, that’s your problem.

    You really got me on that one Axel.
    I feel like a dummy for even considering a different point of view.

    Regarding Ken Saunders post:

    You wrote:
    > “What exactly is open for discussion? Mozilla itself is,
    > but is Firefox?”

    I may be slow… but I failed to glean an answer to your question.

    Regarding my own reaction from all your posts:

    As an old (as in very old) former software developer, I can relate to everyone’s point of view. I can close my eyes and picture some of the nightmares I experienced while trying to communicate clearly to with my clients. I always wrote what I considered to be perfectly intelligent sentences and paragraphs -and- I was often amazed when others took what I said in an entirely different way. It took me a long time to realize that the only thing that is important is ‘what other people perceive my words to mean’. What ‘I think’ is not relevant, unless the other person understands my meaning and intention. And it is my responsibility to do that.
    It is very easy to react defensively, and explain what you meant in a manner that makes the other person feel like an idiot.

    As a long time Firefox user, it really is annoying to do an update and find a little surprise in the package. Screw with my extensions and make them not work? I’m not even interested in the reason (but if you send me your home address…;-)

    Best Regards to all,

    DCStrain
    Minneapolis, Minnesota
    USA
    [personal comments welcome]
    [dennisstrain–aatt–emesen–period–com]