Chrome won

Disclaimer: I worked for 7 years at Mozilla and was Mozilla’s Chief Technology Officer before leaving 2 years ago to found an embedded AI startup.

Mozilla published a blog post two days ago highlighting its efforts to make the Desktop Firefox browser competitive again. I used to closely follow the browser market but haven’t looked in a few years, so I figured it’s time to look at some numbers:

alldevices

The chart above shows the percentage market share of the 4 major browsers over the last 6 years, across all devices. The data is from StatCounter and you can argue that the data is biased in a bunch of different ways, but at the macro level it’s safe to say that Chrome is eating the browser market, and everyone else except Safari is getting obliterated.

Trend

I tried a couple different ways to plot a trendline and an exponential fit seems to work best. This aligns pretty well with theories around the explosive diffusion of innovation, and the slow decline of legacy technologies. If the 6 year trend holds, IE should be pretty much dead in 2 or 3 years. Firefox is not faring much better, unfortunately, and is headed towards a 2-3% market share. For both IE and Firefox these low market share numbers further accelerate the decline because Web authors don’t test for browsers with a small market share. Broken content makes users switch browsers, which causes more users to depart. A vicious cycle.

Chrome and Safari don’t fit as well as IE and Firefox. The explanation for Chrome is likely that the market share is so large that Chrome is running out of users to acquire. Some people are stuck on old operating systems that don’t support Chrome. Safari’s recent growth is underperforming its trend most likely because iOS device growth has slowed.

Desktop market share

Looking at all devices blends mobile and desktop market shares, which can be misleading. Safari/iOS is dominant on mobile whereas on Desktop Safari has a very small share. Firefox in turn is essentially not present on mobile. So let’s look at the Desktop numbers only.

desktop

The Desktop-only graph unfortunately doesn’t predict a different fate for IE and Firefox either. The overall desktop PC market is growing slightly (most sales are replacement PCs, but new users are added as well). Despite an expanding market both IE and Firefox are declining unsustainably.

Adding users?

Eric mentioned in the blog post that Firefox added users last year. The relative Firefox market share declined from 16% to 14.85% during that period. For comparison, Safari Desktop is relatively flat, which likely means Safari market share is keeping up with the (slow) growth of the PC/Laptop market. Two possible theories are that Eric meant in his blog post that browser installs were added. People often re-install the browser on a new machine, which could be called an “added user”, but it comes usually at the expense of the previous machine becoming disused. It’s also possible that the absolute daily active user count has indeed increased due to the growth of the PC/laptop market, despite the steep decline in relative market share. Firefox ADUs aren’t public so it’s hard to tell.

From these graphs it’s pretty clear that Firefox is not going anywhere. That means that the esteemed Fox will be around for many many years, albeit with an ever diminishing market share. It also, unfortunately, means that a turnaround is all but impossible.

With a CEO transition about 3 years ago there was a major strategic shift at Mozilla to re-focus efforts on Firefox and thus the Desktop. Prior to 2014 Mozilla heavily invested in building a Mobile OS to compete with Android: Firefox OS. I started the Firefox OS project and brought it to scale. While we made quite a splash and sold several million devices, in the end we were a bit too late and we didn’t manage to catch up with Android’s explosive growth. Mozilla’s strategic rationale for building Firefox OS was often misunderstood. Mozilla’s founding mission was to build the Web by building a browser. Mobile thoroughly disrupted this mission. On mobile browsers are much less relevant–even more so third party mobile browsers. On mobile browsers are a feature of the Facebook and Twitter apps, not a product. To influence the Web on mobile, Mozilla had to build a whole stack with the Web at its core. Building mobile browsers (Firefox Android) or browser-like apps (Firefox Focus) is unlikely to capture a meaningful share of use cases. Both Firefox for Android and Firefox Focus have a market share close to 0%.

The strategic shift in 2014, back to Firefox, and with that back to Desktop, was significant for Mozilla. As Eric describes in his article, a lot of amazing technical work has gone into Firefox for Desktop the last years. The Desktop-focused teams were expanded, and mobile-focused efforts curtailed. Firefox Desktop today is technically competitive with Chrome Desktop in many areas, and even better than Chrome in some. Unfortunately, looking at the graphs, none of this has had any effect on market trends. Browsers are a commodity product. They all pretty much look the same and feel the same. All browsers work pretty well, and being slightly faster or using slightly less memory is unlikely to sway users. If even Eric–who heads Mozilla’s marketing team–uses Chrome every day as he mentioned in the first sentence, it’s not surprising that almost 65% of desktop users are doing the same.

What does this mean for the Web?

I started Firefox OS in 2011 because already back then I was convinced that desktops and browsers were dead. Not immediately–here we are 6 years later and both are still around–but both are legacy technologies that are not particularly influential going forward. I don’t think there will be a new browser war where Firefox or some other competitor re-captures market share from Chrome. It’s like launching a new and improved horse in the year 2017. We all drive cars now. Some people still use horses, and there is value to horses, but technology has moved on when it comes to transportation.

Does this mean Google owns the Web if they own Chrome? No. Absolutely not. Browsers are what the Web looked like in the first decades of the Internet. Mobile disrupted the Web, but the Web embraced mobile and at the heart of most apps beats a lot of JavaScript and HTTPS and REST these days. The future Web will look yet again completely different. Much will survive, and some parts of it will get disrupted. I left Mozilla because I became curious what the Web looks like once it consists predominantly of devices instead of desktops and mobile phones. At Silk we created an IoT platform built around open Web technologies such as JavaScript, and we do a lot of work around democratizing data ownership through embedding AI in devices instead of sending everything to the cloud.

So while Google won the browser wars, they haven’t won the Web. To stick with the transportation metaphor: Google makes the best horses in the world and they clearly won the horse race. I just don’t think that race matters much going forward.

Update: A lot of good comments in a HackerNews thread here. My favorite was this one: “Mozilla won the browser war. Firefox lost the browser fight. But there’s many wars left to fight, and I hope Mozilla dives into a new one.” Couldn’t agree more.

95 thoughts on “Chrome won

  1. Sad but true article.
    Thanks for your efforts with Firefox OS.
    I still believe that there’s room, a need, and desire for it as much as there is for Firefox for desktops

    • Firefox OS was dropped pretty much when I found it interesting for me as a developer and power user, with os-wide webextensions. It could even have got pretty awesome thanks to containers.
      There is a new revolution that has already started, and everything to define. Firefox is part of it ( but more as spend 2 min in it then come back), but Firefox OS would make a lot more sense in the long run : 6dof VR( +AR/MR ).
      It is here to stay, and with low cost new devices (mobile and desktop ones) which is starting to appear ( e.g a mobile 6dof headset+controllers by htc for japan yesterday )
      My 2 cents…

      • The market in the end judged Firefox OS. We couldn’t catch Android’s growth curve. I am glad we tried. I continue to be pretty hopeful for WebVR. Maybe Mozilla can find a hardware partner for it.

    • Developing Firefox OS for years took a lot of resources away from Firefox though, which is in bad need of help both technologically and design-wise.

      • I think the graph disproves that mostly. 2011-2014 was focused on Firefox OS. 2014 the focus moved back to Desktop with a CEO and strategy shift, and Firefox OS was shut down. The curve looks the same. The best interpretation I can offer is that being slightly worse in a few ways than Chrome (but still mostly good enough) and being mostly competitive and in some areas a bit better (but not massively better in any way) don’t make a difference to user acquisition and retention. Firefox is declining just the same. To really affect that curve you need absurd marketing budget (Google’s owned media alone is probably worth billions a year), or you have to have a massively better product (it’s a commodity market, not likely to happen). In other words that curve is what it is. Dramatic changes in the environment aside it’ll be hard to bend that curve.

  2. The main reason why I switch from Firefox to Chrome was stability. About a year ago when I tried to switch back Firefox would start hanging and would require me to restart the browser. I had posted a question about Firefox being unresponsive and the responder’s suggested it was due to the addons/extensions, which were Camelizer, Roboform, and a couple other ones. One person that responded suggested it was due to the Roboform and to disable it (and Camelize). I had disabled all addons/extensions, but it was still freezing. So, after trying for a month I finally had enough and switched back to Chrome.

    If I can’t use a few extensions that I really need just to make Firefox stable (or alleged to be), then I can’t use it.

  3. To be honest, I don’t think Mozilla has been helping this much. It has been my perception that Mozilla has been following a trajectory similar to the GNOME 2.x->3.x transition for a while. (Disillusioning their entrenched user base while trying to chase some dream user base which may never materialize.)

    Google has given Chrome every advantage when it comes to “commodity users”, which leaves Firefox serving niche markets like “power users” and “users who value their privacy”… and, while things like AMO’s rules regarding analytics in extensions still help the latter, the former will be severely hindered, given the decisions I’ve seen on various WebExtension API requests on Bugzilla.

    In fact, if an embeddable Blink widget like QWebEngine decides to start offering ready-made support for WebExtensions, I’m going to immediately slap together my own browser frontend in something like PyQt rather than using any of the existing ones.

    In the mean time, I’ll be sticking to ESR until 52 drops off it while I evaluate which of SeaMonkey, Pale Moon, a patched Chromium, or Firefox Nightly would provide the closest equivalent in a form I can deploy to people like my mother.

    • Stephan, with all due respect, I am unable to understand the message you tried to get across. I am a developer myself but could not understand why WebExtensions are a bad thing. At least compared to XUL. From https://arewewebextensionsyet.com/ I can see that many methods that can be potentially misused are not implemented.

      • There are clearly some additional extension points that developers need but that mozilla won’t supply (fs access that doesn’t involve a db: https://bugzilla.mozilla.org/show_bug.cgi?id=1246236 or the tabs management api which looks to be put off until after 57 is released which means that users of current tab management add-ons will have an interesting transition come 57). I understand that the api needs to be minimal, but some of the truly useful ff extensions just aren’t possible given the current positions of the maintainers. If reminds me of that saying regarding Open Office and Microsoft Office: OO can do 90% of what most users require, but it’s that last 10% (which probably encompasses all of the missing features from MO) that no one can agree upon which prevents a user from being able to move to OO. In OO’s case (well, LO, now), its been a slog, and you have to determine what makes the most sense to work on (that’s assuming you have developers willing to work on arbitrary features, to begin with). This is similar to the position Mozilla is in with the difference being they have a much larger dedicated group of developers, and a self-imposed, arbitrary deprecation deadline.

      • I don’t think the concept of WebExtensions is a bad thing in and of itself. It allow Mozilla to finally fix Firefox’s aging architecture and it gives me an API that doesn’t for me to jump through boilerplate hoops for simple, common tasks.

        The problem is that, in implementing that concept, they’re deciding that various types of extensions are simply “deprecated, no replacement planned” or dropping to the severely crippled state Chrome extensions exist in because they just don’t want addons to be capable of doing things which could confuse lowest-common-denominator users. (That’s a very minimal paraphrase of their justification for recent WONTFIXings of various APIs.)

        (eg. Hiding the tab strip or overriding its scroll-wheel behaviour would confuse my mother if it happened in her browser, so I’m not allowed to do it in my browser unless I want to use the nightly channel as my day-to-day workhorse and bind against the soon-to-be “internal and therefore unstable” legacy APIs.)

      • It’s quite easy to understand, mozilla is making firefox useless to many of the long time supporters and power users, takes a strong stance against freedom and users just to try to be more like chrome.

        As said on mozillazine forums:
        >> Have the ff devs gone complete xx insane?
        > Looks that way. They’re making far-reaching and very short-sighted decisions in a vacuum. It’s not the first time
        > they’ve blundered, but this is the worst example in recent memory.

    • The browser you missed is waterfox[1], basically firefox without much of the nonsense (removed EME, removed spying telemetry, removed data collection, removed pocket, removed sponsored tiles, removed profiling, restored support for unsigned extensions, support of 64-Bit NPAPI plugins,…). Waterfox dev proposed a project to fork firefox before it drops what makes it useful and make a new browser.

      [1]:http://waterfoxproject.org/

  4. > Firefox ADUs aren’t public so it’s hard to tell.

    FYI: I believe you can see the approx. Firefox ADI on this page: https://www.arewestableyet.com
    Desktop
    Nightly Dev Ed Beta Release
    ADI 30k 98k 1.7M 80M
    for Firefox 52 and later. About 82 million. Plus another 8 million on Android gives a total of 90 million. Add an unknown number of people using Firefox < 52.

    • I believe that’s active daily Installs, not Users, but looking at the data over 60 months it went from 92M to 80M. The data is super noisy (and seasonal). I’ll try to pull more samples and plot that.

      • If I remember correctly many, many years ago Asa Dotzler was using the formula ADI * 3 = ADU as the formula for estimating ADU.

    • ADI is a heavily flawed proxy to usage that has no good relation to actual users. It’s a number that Mozilla is trying not to use any more because it is just the number of add-on blocklist requests and it’s know that there are many reasons why not every installation is sending such a request every day anyhow. Mozilla has Telemetry data internally that gives better impressions of actual active installs as well as usage hours, but that is not public because it’s also easily misinterpreted. Actually, it has been argued many times that the numbers you point to should also be hidden from the public as they can so easily be misinterpreted as this subthread shows as well. That said, while I did have access to more data back when I worked for Mozilla, I couldn’t have published the usage data you may want to know about (at least not without review from the press team etc.) and nowadays I do not even have that access so I honestly can’t tell you. All I can tell you is that ADI are know to not represent what you think they do – I was told even their trend may be completely bogus.

      • Telemetry numbers are really really wrong as the recent pulseaudio debacle has demonstrated, it’s also considered spyware and is the first thing to be disabled by privacy minded people.
        Telemetry is a gadget and should not be taken seriously, no decision should be made based on it.

        Mozilla response to disabled telemetry and bad decisions affecting numerous linux users was “do not disable telemetry or face the consequences” which shows how much they care for their users, user feedback and user privacy.

  5. Technically Google/Chrome had won the war when you were still at Mozilla.
    The web browser is more of a commodity at the moment and switching is hard, which means most power users.. probably stay on Chrome.

    Other users, well, to them, Google IS the web – and Google says to get Chrome – the web well you to get Chrome, they get Chrome. On top of that it bundles itself with any popular software (which is why microsoft removed the ability to change browser defaults programmatically by the way).

    Google is getting closer and closer to also win the web war. As stupid as it may sound, though, their search engine sucks more and more due to the army of people trying to sell stuff, spam, etc. (and google likes to link to sellers, too). I sometimes ponder if having a better engine would disrupt them. It’s very hard, though.

    • Yeah, this war was decided many years ago. I disagree that Google is the Web though. Its dominant on today’s Web. It’s less dominant on mobile, thanks to Apple. The future is Amazon Echoes and self driving cars and what not. Not clear Google will dominate there just because Google won the browser wars.

      • Google, Amazon, and Apple will fail when it comes to self driving cars. The car companies will slaughter them.

  6. It’s really concerning that you conclude that it’s *totally ok* if one browser dominates the market.

    Who cares about standards? We’ll just let Google dictate where the web goes. Because putting all your eggs in one for-profit company’s hands seems like the healthy thing.

  7. For me an essential feature of the Web is content and apps running on independent implementations of open standards, a platform not controlled by any one vendor. That’s what I worked for for 15 years. If that goes away then we’re left with competing vendor-owned platforms, and whether they happen to use JS or HTTP etc is just implementation details. Call it “the Web” if you like, but who cares?

    However, I think it’s a mistake to look at trend lines and think they predict the future. I think the really substantial point in what you’ve written here is the possible Web-compat death spiral. Is Firefox in it, or rather, to what degree is it hurting Firefox so far, and what can be done about that if anything?

    I find your whole paragraph starting “I started Firefox OS in 2011 because …” rather confusing. My understanding of the arguments made for FirefoxOS may be different from yours. I recall that during the rise of mobile, people at Mozilla talked about how desktop would become irrelevant, with browser usage on desktop replaced by mobile (so that Mozilla needed major mobile presence to remain viable at all). I think those predictions were invalid, in that desktop usage hasn’t declined much or at all and mobile usage has mostly been additive. Here you seem to be saying that that was the prediction *all along* but that somehow, even though desktop browser usage is as high as it ever was, (as predicted) desktop is no longer relevant. You seem to advance two different reasons for it being irrelevant — that there’s other usage which is greater (but not by an order of magnitude), and that “Chrome won” so no other browser vendor can do anything relevant on the desktop. The former seems like a non sequitur, and the latter definitely doesn’t seem like a prediction that was made, or should have been made, in 2011.

    I think your horse-and-car analogy is inapt. Cars (more generally, combustion engines) really replaced almost all use of horses, so indeed, horses are irrelevant. If we were still using horses just as much today but also using cars even more, I don’t think we’d say horses are irrelevant. A better analogy would be something like the rise of television: would you argue that that made books and movies irrelevant?

    • We failed to make Gecko relevant on mobile. Through that alone Firefox is in a death spiral. On top of that the desktop Web in general will go away, slowly (many years). Transactions are already moving to mobile already though, quickly. I am not claiming desktop usage will disappear completely but it will simply not be super relevant. Most interaction time happens on mobile going forward. It’s always with you, it has better sensors, it has location, it has context. There will always be people (old folks like you and me :) ), and use cases that will remain on desktop, but it’ll be a relatively small share of the minutes spent online. Maybe we are really saying the same thing here though. We are basically adding 7bn mobile users to less than a billion desktop users. I call that “desktop is becoming an irrelevant corner-case for a few people”. You call it “additive”. The new wave of online users is not buying desktops. I think on that part we can agree?

      • Depends on what you mean by “the new wave”. There are lots of people coming online who are mobile-only, agreed. But there are also lots of kids coming online who use desktops or laptops as well as phones. Doing your homework on a phone is still not a great experience.

        Fundamentally I think we have different meanings for the word “relevant”. I’m never going to call a platform that gets a billion person-hours a day “irrelevant”, no matter what else is happening.

        If we went back in time to 2011 Mozilla and said “hey actually desktop usage is going to grow in absolute terms for at least six more years”, I think that would have changed the conversation. For me anyway.

      • For me it wouldn’t have actually, so we just have different perspectives there. I am not saying you are wrong. My viewpoint: back when we both worked on this a billion people hours a day was insane. Today mobile has several billion usage hours a day, and keeps growing further. Long term the much smaller thing (desktop) won’t be relevant. There is already a lot of stuff you can only do on mobile. That’ll keep growing. The world if optimizing for the common case: mobile. I am out of ideas how to win mobile at this point, but there is clearly stuff coming after mobile. VR, IoT, AI, and others. My advice is to focus on those instead of worrying about what happens to the desktop. We can influence a lot of people’s online lives in the future without ever worrying about the desktop again.

        Fun story by the way. Talk to kids about smartphones and homework. I have seen first hand kids doing homework on a phone, resting their elbows on a brand new MBP. Kids grow up with touch. Laptops are unfamiliar to them. Maybe tablets will fill this gap, but keyboards are definitely not something that is intuitive to you if you had a smartphone by age 3 or 4.

      • That’s a good story, but there are also plenty of kids who grew up on touch devices who use keyboards for their homework. Data would be interesting.

        FWIW I think dividing the world up into “desktop” and “mobile” can mislead. The health of Firefox doesn’t depend on the presence of a keyboard. I think the real distinctions are more circumstantial; the current dominant “desktop” platforms, Windows and MacOS, happen to be ones where the built-in browser has limitations, the platform is open enough that a third-party browser is viable, and for historical reasons Web sites targeting large-screen non-touch devices tend to configure themselves in a more standards-friendly way. ChromeOS is changing that; it has the form factor of a desktop platform, but not the third-party browser viability. If Android expands into that space, or Windows and MacOS get locked down a lot more, then that also closes up the window for Firefox.

      • I was going to disagree with you.

        But then..I am replying using a mobile as I got this link on a slack message…while in front of a 27″ iMac….so yes.

      • “Windows and MacOS get locked down a lot more, then that also closes up the window for Firefox.”

        Windows won’t be closed to Firefox or other browsers, even if it became more locked down. The only reason Windows 10 S doesn’t support Chrome or Firefox is because those browsers will never come to those stores.

  8. I chose to move away from The Fox when the Mozilla employees decided that their leader had no right to his own political views. Sayonara, babies. There are choices we have to make in life, some more important than others.

  9. Firefox 57 will leave legacy addons behind, and I think this will trigger a step-function drop in market share. I am not really sure what I will do without Vimperator, but I will be looking at either at the forks or a keyboard friendly browser. The “vim” plugins on Chrome were not nearly as good when I looked last time, and Chrome could not handle my use case of having a few hundred tabs open,

    • It’s a tough spot for Mozilla to be in. The add-on ecosystem needs to be redone to have competitive performance which causes friction, losing users. Difficult place to be in. I know the team that does this quite well. They are doing their best to keep frequently used add-ons working.

      • If that were purely the case, I wouldn’t mind so much, but I’ve been monitoring the situation for migrating my addon loadout and I’m running into more and more situations where they’re either reticent or flat out WONTFIXing APIs which are non-negotiable for implementing certain types of extensions. For example:

        1. I’m very worried that, for any addon which needs disk I/O more involved than programmatically-generated downloads, I’m going to have to write a local HTTP daemon which reinvents XULConnect. (Because it’s just too much hassle to install a native messaging handler, while a localhost HTTP daemon with popup permission prompts would be easy to share between multiple addons.)

        2. They just WONTFIXed the ability to override tab strip scroll wheel events, when less then a year ago, they said that making Firefox’s tab strip scroll-wheel behaviour match every other Linux application including Chrome was a task better suited to an addon in the bug about offering it internally.

    • The Mozilla people have looked at the number of users whose addons will stop functioning, and it’s actually pretty small.

      You could argue that those users are especially influential so there will be a flow-on effect, but nobody really knows.

  10. what does that mean “an exponential worked best”? highest R^2? that’s a pretty strong model choice.. ask Malthas. The line for chrome looks a lot more concave than convex.

    • Yeah Chrome didn’t fit well at all (as I sad in the post). If you have a better suggestion make a graph and I’ll update it in the story. At the high level most analyses will look the same I think.

  11. I think the general sentiment that things that aren’t on a huge growth curve aren’t relevant or worth doing is very unhealthy–especially for the purpose of drawing conclusions about what Mozilla should do.

    What should the conclusion have been from growth graphs back when IE was growing? Everyone else just give up?

    The growth of phones doesn’t make desktop irrelevant. Phones can’t replace the desktop paradigm for many tasks, so desktops will stick around. When they stick around, things are better if desktop browsers stay in good health.

    As for mobile, people do browse the Web a lot of mobile. Maybe the numbers for Facebook and Twitter apps are even greater in terms of minutes of use, but that doesn’t make the Web on mobile irrelevant or not worth caring about.

    Whether on mobile or desktop, we’d all be worse off if the Web becomes less healthy due to neglect arising from being thought as not growing enough to be cool or being left to only to operating-system-bundled browsers. (The low switching cost of browsers due to not having to switch devices keeps the browser space competitive even if also tough for Mozilla when Mozilla doesn’t have an OS as an anchor.)

    Sure, non-Web things (IoT and other) will exist and individual people may be a bit tired of doing Web and go do those other things. That doesn’t mean that the Web isn’t worth caring about anymore or that it doesn’t continue to be important for the health of the Web for Mozilla to be there with an independent engine–doing what Mozilla has always done without pivots to new shiny.

    When individuals grow tired of Mozilla or the Web, going to do some other thing elsewhere for a change is cool, but I think spreading defeatism about what those of us who stay do is not cool.

    • I think those are very different times. Mozilla fought IE at a time when the (desktop) Web was exploding. Mozilla helped that Web blossom and grow and established its share. Today the desktop Web is not growing much, and capturing market share is a very different problem. And I didn’t say that I don’t care about the Web. To the contrary. I don’t know how to fix desktop or mobile. It’s what it is. Chrome won. There is no way I know of to get market share back there. My point is rather I don’t think we have to. New stuff is clearly on the horizon. If we win that, desktop (and at some distant future point even today’s mobile) won’t matter as much.

    • Servo and Rust are amazing technologies and a decade ahead of the state of the art rendering engines. I have a hard time seeing how that would change Desktop or Mobile browsing. Hardware has become so fast its good enough. If my browser was X times as fast I am not sure I would notice for general use cases. VR may be different. 8K 16K 32K all still make a qualitative difference for VR. Those are insane pixel counts. If WebVR takes off it’ll take Servo to render it at peak fidelity. There are probably lots of other forward looking use cases like VR. That would be my recommendation for Servo. Look forward to VR, maybe streaming gaming as well, and other upcoming use cases of the Web. Don’t look back to Desktop, and likely not Mobile browsing either. Better fights to be won (and winnable).

  12. Firefox with duckduckgo is awesome! I’ll never go back. Chrome is bulky, cumbersome, lacking in security and full of trackers. My privacy is respected and I get everything I need from Firefox. It is customizable and lightweight. I guess Android and gmail helped to get people hooked on Google and on Chrome. Security & privacy are becoming massive issues, this could save Firefox. I realize I’m in the minority regarding my views on Google and Chrome, but it doesn’t mean I’m wrong. Step out of your comfort zone, ditch all your google services, there’s a whole new world waiting for you out there.

    • I always see people complain about Chrome’s RAM usage, and honestly it annoys me. Are you using the RAM for anything else? Obviously not, or Chrome wouldn’t have taken it. In fact the only times I’ve ever noticed Chrome using more than a couple hundred megabytes is when my cookieclicker windows gets a RAM leak. There are plenty of reasons to dislike chrome, but resource usage shouldn’t be one.

      • Actually, yes, I *am* using that RAM for something else. I bought 16GiB of RAM when 4GiB was the norm because I do a lot of multitasking and parallel operations while I develop.

        While I can’t currently afford to upgrade, I do often have to close down Firefox or Chrome tabs because my system is starting to thrash.

        (Thanks to a $7 passive DisplayPort->DVI adapter, a $3 HDMI->DVI cable, and cheap, used DVI monitors, I have a three-monitor spread which I use for things like “Vim and log output on one monitor, Firefox with auto-reload on a second, Chrome with auto-reload on the third for rapid iteration, and a couple of modern.ie testing VMs in the background ready for a click and an F5”.)

  13. Thanks for this insightful post.

    I agree Chrome has won and Mobile disrupted the web and ultimately the browser. From personal experience, chrome on the desktop dominates because of this vast ecosystem of Google products(Search, Gmail, Drive, Calendar) that literary own my life and make it(my life) easy across all platforms and various devices. Chrome provides the underlying infrastructure that makes navigation of its products easy. On contrary Firefox only has a browser which at this point alone offers little value to me.

    On mobile, the social and visual internet is more predominant. This is vastly supplied not by browsers including the predominant chrome, but by just a few social giants; Facebook, Twitter, Whatsapp, Telegram, Gmail. You could say that Facebook app is the browser of mobile because it does on mobile best what browsers do on desktop. A new wave of chatbots integrated with the Facebook ecosystem of products is creating a whole new mobile web in which bots are to mobile what websites are to browsers on the desktop. Smart [voice] Assistants powered by AI are now also creating a whole new invisible computing interface in which buttons and icons that defined the user experience in browsers are obsolete in mobile space.

    I guess Mozilla has to rethink has it considers to be the web today in a ways that are different from the earlier desktop pc era.

    • “Chrome provides the underlying infrastructure that makes navigation of its products easy.”
      Huh? I use Google search, Mail, Inbox, Calendar, Maps, and Drive in tabs in Firefox and they all work fine. If there really is improved navigation of them in Chrome, I’ve never noticed it. Are you sure all those “try Chrome it’s better” ads didn’t skew your perception?

    • “Chrome provides the underlying infrastructure that makes navigation of its products easy.”
      Huh? I use Google search, Mail, Inbox, Calendar, Maps, and Drive in Firefox tabs and don’t notice any difference in “navigation ease” compared with chromium.

      What exactly is better about Google’s’ web apps in Chrome, besides the absence of nagging “Try Chrome” popups?

    • Yeah. Someone on HN said Mozilla won and Firefox lost. Well put. I don’t think Chrome will lose to another browser. But it’ll lose to something. VR. AR. Assistants. Something. Time to swing for those things.

  14. Very brave perspective on the future but I think it’s a little too brave, as in, many predictions won’t happen in a relevant time range, aka my lifetime or at least decades. Might as well worry about the existence of human kind if population continues to grow at this pace.

  15. If Firefox wants to get back market share the solution is simple. Browser with either ad blocking or adblock support (api/extensions) on mobile.
    There’s nothing good out there and ads are really intrusive on mobile especially due to smaller screen real estate.
    That would get you a bunch of installs.

  16. Pingback: Chrome won | Andreas Gal | Artificia Intelligence

    • The foundations for distributed apps are 100+ web frameworks, mostly open source, developed by 40+ different companies in 10+ different languages. How would a 101st from Mozilla make any difference? (And actually, the 102-107 web frameworks are being developed in Mozilla’s Rust language).

      Mozilla can only advocate like hell (along with the other also-ran, Microsoft) that those foundations use web standards rather than Chrome-only hacks.

      • You misunderstand me. My fault. Groups like @Blockstack and @ethereumproject are defining development models for the decentralized Internet and distributed apps. In my mind Mozilla absolutely should have a voice in setting standards. At the very least they should be ensuring that Firefox is a first class citizen in those ecosystems. Why is Mist (ethereum) even a thing? Why is Blockstack defaulting to Google Chrome? Granted for the Blockstack part that’s easily changeable, since they’re just using BrowserSync. But still…

  17. Pingback: Former Mozilla CTO: Chrome Won The Browser Wars

  18. Clearly Chrome is winning and has won. But your curve-fitting undermines the message: the Chrome trend line does not correspond to the actual data in any realistic way: it should be concave down-and-to-the-right, approaching a steady state (horizontal) about 75% to 80%. This flaw incites suspicion among readers, including me.

      • Still would be valid to add, since its still regarded as a major browser (and it only just started, so it is to be expected). But you not only ruled it out of the graph, you ruled it out of the whole article. IE is discontinued so its pretty simple to say that it won’t get more users because thats what Edge is for.

      • Edge may gain share. I am not disputing that. It’s a solid browser. It could be like iOS: hard to displace on machines Microsoft controls.

  19. What does this data look like if you also toss Chromebooks, since they’re basically the equivalent of an iDevice as far as browser choice? I don’t imagine it would be drastically different, but it only makes sense to narrow those out if you’re not looking at mobile devices either.

  20. I’m a firefox user for one reason. Chrome runs like crap on macOS. Firefox is much more lightweight and doesn’t lag on large websites like espn.

  21. Pingback: Chrome ha ganado la batalla, pero no la (verdadera) guerra | JanyoBytes

  22. Pingback: Chrome ha ganado la batalla, pero no la (verdadera) guerra | Android 3G - ¡Estamos conectados con la Tecnología Android!.

  23. Pingback: Chrome ha ganado la batalla, pero no la (verdadera) guerra – Olrait Diario

  24. Pingback: Chrome ha ganado la batalla, pero no la (verdadera) guerra | ISO Móvil

  25. I agree that the statistics are bad for Firefox, but I feel the need to point out a few points that are important to me.

    Windows itself is losing to Linux/Android and Darwin/iOS. The “distros” that are beating Windows are walled gardens with corporate spyware, of which Chrome has a non-zero share.

    The time may well come where corporate-controlled Linux and Darwin are themselves beaten by open-source versions. Should such a situation arise, Firefox is well-positioned – it is the default browser on RedHat/CentOS/Oracle Linux and is more easily available on similar platforms than Chrom[ium].

    It should also be pointed out that Chrome forbids extensions on Android that are allowed on Windows. Firefox does not suffer from such arbitrary rules.

    Chrome is not good for control, privacy, and accountability. Firefox is, and that may count more in the future than it does now.

  26. There is no way to say what I have to say without sounding like I am trying to insult you (because of your intimate involvement with Firefox) but I swear it is not meant that way.
    1) I don’t agree with you that desktop is dead as a browsing platform compared with mobile/whatever else the future brings. I am sure on a long enough timeline you are right (see Fight Club) but I don’t think it’s wise to just cede the desktop landscape to Chrome and call it a day. Desktop browsing isn’t going anywhere until people can directly replace that experience with something as easy and efficient. I’m sorry, but just because a few kids do (some) of their HW on a mobile device doesn’t mean that desktops are dead. They still offer much larger screen real estate, easier controlling peripherals (mouse/keyboard/drawing tablet/etc), and they offer much more performance compared to most mobile devices at the same general price point. Most of these advantages are because of the form factor and innate desktop-iness of desktops, so those natural advantages won’t disappear unless mobile devices get some innovative new technology in a hurry. I’m not saying mobile stuff won’t project a huge screen somewhere, be compatible with more efficient UI peripherals, and potentially bridge some of the performance gap with stationary home computers……but it’s probably not all happening in the very near future.

    2) Part of the reason Chrome has excelled and Firefox has languished is that Firefox has had either too many wrong directions or no discernable unique direction (take your pick). Firefox staff seemed to be intent on stripping out many of the features that formed the core of what initially drew users to Firefox in the first place (plugins, unique UI, theming, innovative features, etc). Firefox alternately would mention that legacy plugins were to be discontinued, theming would no longer be possible, amazing tab management experiments that were to be “the future” were being nixed, etc. Meanwhile, even an experienced FF user can barely tell whether they are using Chrome of FF when in either browser, FF announced it was potentially moving to a Chromium base, and we all waiting for FF to play catch-up on pretty major features like process isolation. This is a far cry from the FF that made me a loyal convert SO MANY years ago. I still use FF on the desktop because it is so customizable from a plugin standpoint and because it runs stable for me, but even I will admit there isn’t much difference between it and Chrome in those areas anymore (which is sad).

    3) FF is just not as good as Chrome on android. I’m not sure who’s fault this is but it’s true. Mobile Chrome is quite stable. I routinely have more than 100 tabs open and if I open 30+ incognito tabs as well Chrome will slow down and fail somewhat gracefully after a while. Naked browser is equally as crash-proof on androd (maybe more) but the interface is not for everyone. FF on android, by comparison, is not nearly as stable. I still have it on my phone but I almost never use it unless I want to take advantage of some plugin for it. I used to use it more on android ICS and Lollipop but I have almost abandoned it now on Marshmallow in favor of Naked browser and Chrome.

    Things Mozilla should do (IMO):
    A) Fight for desktop browser market-share. People that love FF will come back to it if Mozilla can offer a compelling product. IE is basically dead, Edge is….uh….quirky, Safari is niche, and only Chrome is a compelling figure. How is this an unwinnable fight? For goodness sake, FF used to be “Phoenix”….if it can’t rise from the dead what can?Chrome is ok……but I know a few power users (multiple 100s of tabs open in FF and Chrome at any given time….one person had nearly 1k one time in FF) who have demonstrated that FF is superior to Chrome when it comes to tab and resource management. It is much harder to finesse Chrome to get it stable at that volume of tabs than it is to make FF behave. There is still good in FF. Re-embrace what FF users actually want – customizable UI, plugins to personalize the experience, safety, reliability, CHOICES (something Chrome doesn’t usually excel at).

    B) Make FF look DIFFERENT from Chrome and stop following like middle child with an identity crisis. I should actually be able to tell what browser I’m in without having to think too much about it.

    C) Listen to the damn users of your browser before we ARE all gone. Regardless of the feedback in the forums it seems like FF development continues on a preset course unless the kickback is so strong that it demands an answer. Since Chrome marches to it’s own drum seemingly regardless, I believe this is an area where FF can gain some traction.

    But then again, I’m only a frustrated longtime user of FF who worries that it has completely lost it’s identity. It is equally disappointing to see a former member of the team so readily admit defeat in the desktop arena, call it dead anyway, and refuse to re-engage in what I consider a primary area.
    Perhaps the experts know better, but if you do…..it’s a hopeless future for Firefox IMO.

  27. Apologies for the small spelling miscues and improper usage of “it’s” where “its” should be. It doesn’t seem like I can edit so you’ll have to forgive me.

    I hope FF can regain some purpose in life and start to shine again. I believe there is room for it to challenge the big players, mostly on the (dead) desktop landscape since there is really only one browser to challenge at the moment.
    FF definitely has an uphill battle though, as it seems to have eschewed customization, uniqueness, and innovation for rigidity, mimicry, and playing catch-up.
    I look forward with the hope that Mozilla will at least try……something new and different. FF is currently a shell of its former self and yet I still use it at home.

  28. FWIW, your Chrome trend lines look flawed and therefore it makes it look like your whole post is a Google ad laced with a Silk ad and bashing of Mozilla.
    That said, I also do not agree with the sentiment that desktop usage of browsers is shrinking at all. OTOH, mobile usage is still rising and overtaking desktop usage, even though the usage profiles/patterns differ quite a bit. Tablets don’t really matter but phones do. There is IMHO not much innovation in browsers on any of those form factors any more, as you say, they are a commodity. Mozilla is in a tight spot because the other browsers caught up and are good enough, on both desktop and mobile, additionally, both relevant mobile platforms are anti-competitive in the browser space and either straight-out disallow real alternative browsers or disallow shipping them as default installations and users see no really good reasons to switch to an alternative, esp. for the mobile use cases. I agree that Chrome won (so far) but mostly not because it’s a superior product but because it has the superior amount of ad money, did the better marketing overall, and because Google made Android win and forced Chrome on people as its default browser. I agree with other people posting though that Google is the web nowadays because any new technologies on the web are pushed by Google first or at least faster than by everyone else, including by Mozilla. In addition, Mozilla in recent years did lose its grasp of what its identity is and what its community and enthusiasts want, and a lot of former Mozilla enthusiasts have not been pushing Firefox to their peers any more because they didn’t feel that convinced of it matching their message any more. Mozilla is trying to reconcile some of that, but the rate at which Mozilla loses people with strong ties to the community from staff makes me wonder how well this is going to work – that said, I’m still a community old-timer with an interlude as stuff who still pushes on as a volunteer to both make Mozilla better better and push the message, and there are more, but there’s only so much we can do and former execs posting about how Mozilla lost is not helping.
    And Firefox OS is dead because Mozilla execs did align with Telco partners more than with the Mozilla community, ultimately pushing phones that neither the mass-market nor the enthusiasts really liked. I see the fault mostly with the execs that bent over backwards to satisfy those commercial partners and with the execs that later did not really embrace the switch to cater to enthusiasts that started to happen in the endgame. If web innovation on mobile would have any chance, it would need to be through a platform like FxOS was but there is none left.
    When it comes to IoT, I share your hope that the web plays a role there, and there is a lot of innovation to be done in the general space (as opposed to the desktop or “traditional” mobile spaces), but right now I don’t see how the web plays or can play any major role there outside of devices that expose a graphical screen. JavaScript may play some role, mostly via Node.js, but that’s not the web, it’s just another script programming language. The real web stack of HTML+CSS+JS+DOM etc. IMHO becomes useless when you don’t have a screen, touch or not. The Internet in general is of course a major part of IoT by definition. That said, the spirit of the Mozilla Mission, openness and interoperability (in standards and development), availability to everyone and participatory/decentralized design, people having control over their experiences and free ability for people to make their own choices are definitely in danger by a lot of technologies currently rising in that space and also important to hold up and push for in this space, so I welcome everyone pushing for Mozilla’s values there and for less dominance by large corporations that tend to place profit before those values.

    • Please read the text. I agree that Chrome doesn’t fit well the exponential curve. If you find a better fit let me know, but I doubt it would come to any other conclusion that Chrome is dominating and will continue to grow. It’ll likely flatten eventually as it runs into iOS market share, which is immovable. Everything in between will be crushed. As for Firefox OS, we can argue all day why it didn’t work and what else could have worked, but in the meantime let’s just agree it didn’t work out and now it’s too late to try again.

  29. “…because Web authors don’t test for browsers with a small market share. ”

    Find this to be one of the truest points to this blog post, duplicating interaction tests to use the Ff driver kit or elsewise often feels like more trouble than it’s worth to ship web product. Agree with others here, it’s a sad truth.

  30. Desktops aren’t going anywhere. The person who started saying that they’ll die was probably just someone heavily invested in mobile devices.
    The sites and services that people use and browse through on their mobile devices are built on desktops (says captain obvious).

    Research, advancement, management, implementation (and so on) for science, movies, music, medicine, technology, military, space, and others are, and will continue to be done on desktops.

    We’re not going to see a doctor performing a robot assisted operation using a device that someone plays Candy Crush on.

    • That medical device will look like a tablet on the inside, not like a PC. Just like ATMs all started look like a PC on the inside 20 years ago. Supply chain dominance. We make a lot more mobile devices these days than PCs. A lot lot.

      • “That medical device will look like a tablet on the inside, not like a PC.”

        I don’t doubt that, but that medical device and the software for it will be designed, tested, etc, using a PC.

        I don’t doubt the capabilities of mobile devices at all, but they are more for managing things than they are for building things.

        “We make a lot more mobile devices these days than PCs. A lot lot”

        I get that and I get why. My point was that desktops are not going away any time soon but many people have been declaring that they are for a long time, and some even said that desktops were dead several years ago. And in my opinion, software development for desktops suffered because of that.

      • I agree. PCs will be a around for a very long time but it’s a niece product. The vast majority of eyeball time will be on mobile devices.

  31. I think there’s a role for a browser-based ecosystem that doesn’t involve sharing your data with a privacy-sucking third-party social web site. I don’t want Google to know my todo list, my personal notes, or my reminders. But I do want to access this data in a browser and sync across devices. I hope the experience of developing local apps for Firefox OS assists in the development of such add-ons. If they are cross-browser all the better, but if they work with Firefox Sync they would be more compelling in Firefox. (FWIW the closest I’ve found for my needs is TiddlyWiki and variants.)

  32. Pingback: #freesw wins here. That’s why #chrome did NOT win and I tell people… | Dr. Roy Schestowitz (罗伊)

  33. On my desktop I exclusively use Firefox for browsing and will do that until possible, but whenever I test websites with Chrome I can clearly see why Firefox’s desktop market share is falling: Chrome is insanely responsive on slow machines. Loading, rendering, reflowing simply feel faster with Chrome. Complex and large websites cause Firefox to become unresponsive, Chrome doesn’t care. The benchmarks might not show this, but my experience is different. I have high hopes for Servo, but it looks like we are years away from the switch :(.

  34. Pingback: Voice of the Masses: What’s your favourite web browser?

  35. I still use Firefox on one of my OSes because I prefer it’s download platform, But everywhere else I switched from Firefox to Chrome because of two things : Firefox took much more time to open, and Chrome was always the first to support new web features. Since I am a web based software developer in an environment where I can choose on which platform my software is run, I found Chrome’s leading edge on new web standards very useful.

  36. Pingback: Let’s increase Firefox Market share – Md Shahbaz Alam

  37. I still use Firefox on one of my OSes because I prefer it’s download platform, But everywhere else I switched from Firefox to Chrome because of two things : Firefox took much more time to open, and Chrome was always the first to support new web features.
    The problem is that, in implementing that concept, they’re deciding that various types of extensions are simply “deprecated, no replacement planned” or dropping to the severely crippled state Chrome extensions exist in because they just don’t want addons to be capable of doing things which could confuse lowest-common-denominator users.

    • “The problem is that, in implementing that concept, they’re deciding that various types of extensions are simply “deprecated, no replacement planned” or dropping to the severely crippled state Chrome extensions exist in because they just don’t want addons to be capable of doing things which could confuse lowest-common-denominator users.”

      It’s a bit more involved than that.

      Chrome supports an extension capability, but Chrome extensions are pure JavaScript. Firefox is going in that direction – any extension available on the Mozilla Add-on site that says “No Restart required” is also pure JavaScript. Microsoft Edge is in the early stages of supporting extensions, and those are also pure JavaScript. One goal for the browser developers is cross-browser extensions that can be installed in Chrome, Edge, or Firefox.

      These will all be pure JavaScript, and will be limited to what JavaScript is permitted to do. *That* has been steadily lessening. Among other things, you must jump through various hoops (on Firefox, at least) to do anything in an extension that affects the local file system. JavaScript increasingly isn’t allowed to do that. But affecting the local file system is exactly what an assortment of widely used Firefox add-ons must to to perform the tasks they are used for.

      This is intended to provide better security, but with the side-effect of reducing capability. As an end-user, I want the ability to say “Yes, I understand this may be a security risk. I know what I’m doing, and choose to accept that. Let me do it!” Since the various devs don’t seem to be interested in what end users want,. I fear that won’t happen.

  38. Pingback: Firefox marketshare revisited | Andreas Gal

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s