The new browser wars: an analysis

A billion years ago I worked on web browsers. I’ve written about them before, and got myself into trouble here and there for what I’ve said. I get asked often what I think about Chrome, or an innovation analysis of Opera, or this or that, and it’s a good time to look at what’s been going on.

Recent data showed that IE has fallen to 59%, the same share of the market it had when I worked on IE 4.0 in 1999.

Here’s my analysis:

  • All data is biased. There is no single count on market share, so never allow yourself to get the final word from one sample. Browsers, like cars, reflect demographic differences. The browser data from slashdot.org will be very different from wsj.com, because of age, income, gender, profession and other differences that are not representative of the total population.
  • A large percent of browsers users do not choose their browser.  One play Microsoft made in IE4/IE5 was to invest heavily in IT deployment of IE (See IEAK). They made it very easy for large companies to rollout IE across 1,000 of desktops, including libraries, university’s and other places that control many desktops.  This is likely an anchor of Microsoft’s market share as I don’t think anyone else does as much for them. This user group is notoriously slow to upgrade or change.  It is hardest for Chrome, Firefox or Opera to penetrate this usage base. Big shops bet more than just the browser on IE: they bet lots of web apps and internal tools. To switch requires rebuilding that infrastructure, making the choice much larger than just browser installs. The browser is free but there’s much resting on the decision that isn’t.
  • Firefox’s growth has flattened in part because of the above. Most of the people who care what browser they use and have the choice have already chosen. The next wave of growth has to be fueled at reaching out beyond Firefox’s core base of younger, more tech-savvy people (or people who want to pretend to be young and savvy). Chrome has likely been the Ralph-Nader of browsers, stealing some of Firefox’s thunder. Opera is still the same creative but lonely story , a very inventive product that most people rarely hear about here in the U.S.
  • We forget most people don’t care much about browsers.  If you read this article, and have seen browser stats in the last month, you are a browser geek like me. Few others on this planet care much. If they’re my age or older, and the web isn’t a key part of their social life, they’ll use what they have and unless it explodes or makes them cry, they won’t think about it. At WordCamp SF this weekend, I heard IE referred to as “the browser for old people”. A browser is one of the most passive kinds of software ever invented: it’s a platform for the websites not for itself – the browser UI occupies maybe 10% of the screen at all times making it forgettable by design: if its working well you shouldn’t notice it. In a great post on pingdom, they studied the upgrade rates for each browser, and IE users upgrade the least, or IE does the worst job at convincing its users to upgrade.
  • HTML5 will not save us. The world is confusing to most people in part because they’ve never attended a standards body meeting. It’s a bloody, messy, clumsy, slow and frustrating process for everyone involved, and although it is necessary and I believe in standards and I’m glad we have the W3C, the process is always problematic and painful. HTML5 will be progress, for sure, but there will always be the same challenges of sorting out what Chrome did, but IE didn’t, and what FF did right but Opera or Safari did wrong, or just differently. Wise standards bodies depend on implementations to validate their standards, and as the different browsers are competitors, it’s never a straightforward process – it’s a huge unavoidable compromise – meaning the results are never as straightforward as we wish either.
  • Since the killer feature is dead, the battleground is on core (speed, security, reliability) which are not Microsoft’s strengths (certainly in perception, if not reality).  Microsoft is still dominated by the annual release cycle, with big marketing pushes around the new features in each release. But Chrome has made a different bet, aimed both at IE’s weaknesses (big, heavy releases) and Google’s strengths (speed, less legacy code, web-app focused). They are pushing an argument, supported by the vibe of the web – catching up on features can be done, but catching up on perf and security is much harder (even if we’re just talking perceptions of perf and security). Firefox is curiously in the middle: with the common bevy of tabs and extensions, it’s not a spry little browser anymore, but instead a happy compromise between Chrome and IE. The rub is they are now being attacked from both sides, above and below, and they’re no longer the lightweight alternative.

Stats I want to see:

  • What percentage of browser users have at least one plugin? I mean one that they chose to add themselves. I bet this number isn’t as high as we all think it is, but I’ve never even seen an attempt to document this. It might be a useful comparison between browsers: both the number of available plugins, and their frequency of use, per browser.
  • What percentage of installed browsers could be changed by the user? I’ve never seen this stat either, but it’s key. If there are 50 million web browsers open right now, how many of them are operated by someone who is allowed to change it? And knows how? It would indicate to both Microsoft and Firefox exactly what they need to protect or go after.

Personally I’m still happy with Firefox. I gave Chrome a spin when it was first released, and was pleased, but not enough to switch. The handful of FF plugins I use keep me tied to it, in what is a a  kind of meta-killer feature: plugin addiction. I’m locked in to the plugins I have and they’re browser specific, forcing me, post facto, to stay with the browser I have.

6 Responses to “The new browser wars: an analysis”

  1. Mark Connolly

    Your analysis reads like you’re thinking in terms of the traditional web (for lack of a better term), though I may be mis-reading you. How, if at all, do you see the mobile web affecting browser development and adoption?

    For me, if the mobile web is factored in then Webkit-based browsers (Safari and Chrome) become more than the spoiler amongst browser choices. I’m curious to know whether mobile browser usage has any affect on what traditional web browser a user chooses.

    Reply
  2. Michał Paluchowski

    The part above which I would particularly underline is the users mostly not caring about which browser their using – that’s VERY TRUE. I got reminded of this lately when teaching an elderly aunt some computer basics (she has her first ever computer for a few weeks now.)

    I did eventually choose Chrome for her just because it had the least UI elements on screen, so also the least possible points of confusion. She switched form Internet Explorer, which she started out with, easily after being ensured that Chrome will allow her to achieve exactly the same things.

    A novice like that may be an extreme case, but most users are really not much further into technology than my aunt. All they need their browsers to have is an address bar, a ‘back’ button and a bookmark storage.

    Reply
  3. Asa Dotzler

    I’d guess that very few people have installed a traditional Plug-in (like Flash or Java) and that the ubiquity of Flash is thanks to it being bundled with Windows XP and then bundled by pretty much every OEM after that. Java shipped with Windows and through OEM channels and it’s probably the second most used plug-in. Other plug-ins which didn’t ship with Windows have very poor penetration.

    Firefox Extensions are a bit different from traditional plug-ins and I think have quite a bit more traction. We estimate that about 1/3rd of Firefox users have at least one extension installed. http://bit.ly/yd3OB With nearly 400M users worldwide today, that’s a non-trivial number. I’d bet that extension use in Chrome is also substantial. Google could certainly tell us, with very specific numbers since they track individual users pretty closely.

    Reply
  4. DelC

    One interesting point that you don’t make is what impact the Browser Ballot that has been pushed in the EU might have on this. Although this issue crystallised around Windows 7, it has also been pushed out to XP. As you can see from the links below Opera have been claiming that it has helped them.

    http://microsoftontheissues.com/cs/blogs/mscorp/archive/2009/07/31/windows-7-and-browser-choice-in-europe.aspx

    http://news.bbc.co.uk/1/hi/technology/8574883.stm

    Reply
  5. Kerry

    In terms of numbers of ad-ins, there is also the issue of ‘standard’ plugins, eg Flash, which would need to be discounted. I mean, seriously, how many Firefox users don’t have Flash support installed?

    Reply
  6. Relgoshan

    Opera has been so very busy with engineering, that it has not bothered much with merketing. This has begun to change. There are new satellite offices in Australia and Russia, new projects and investments. Opera seems to only be able to “bet large” and hope for the best.

    It’s not to say that Opera couldn’t be doing some things a little better, but it works on every PC I own, and I always find my needed features lurking under the hood.

    Please to compare the current state of “opera:config” and “opera:cache” against the Firefox equivalents.

    Reply

Leave a Reply

* Required