I would like to know if usability is different across different Internet browsers and if this is caused by varying demographics. Below, are some interrelated questions addressing this issue:

  1. Does age, gender, or profession affect a user's choice of browser?
  2. Are users of certain browsers more tech savvy and do they muddle through usability problems more easily than users of the remaining browsers?
  3. When conducting usability tests, will forcing the participants to use a browser that is not their preferred browser corrupt the findings of the test?
  • 1
    Depends whether you are further able to categorise it by home or work use as many organisations IT departments determine the applications available to users, type of browser for example - which is why Microsoft have only jrecently been able to announce the death of IE6 bbc.co.uk/news/technology-16408850
    – user11427
    Commented Feb 4, 2012 at 21:41
  • 2
    @Kingsley IT departments that require IE 6 are the spawn of satan.
    – Zelda
    Commented Feb 4, 2012 at 23:37
  • 1
    @BenBrocka true, but in this economy users can't always flee satan. Commented Feb 6, 2012 at 21:43
  • @MonicaCellio: Still, that is an answer to the question. The demographics of IE6 users are indeed unique.
    – MSalters
    Commented Feb 7, 2012 at 15:16

4 Answers 4

  1. It's consistently been shown that Firefox is more popular with younger demographics than IE and Chrome is the most popular browser amongst young and tech savvy users. IE is still most used by older users. Don't take this to mean you can assume your IE users are old or that your Chrome users are young, however.

  2. Historically it's certainly been the case that less tech-savvy users are less likely to install different browsers, thus Internet Explorer and Safari are common on the PCs and Macs of less tech-savvy people. Chrome was shown to be a bit more popular for users that spent more time on the internet but the survey is from 2010; I haven't been able to find much recent information broken down by demographics.

    However since Firefox and Chrome have grown (and have been very effective at simplifying the install process) the browser wars are less tied to tech savviness; it's no longer an indication of tech-savvy to have a different browser installed.

    Don't just assume that because you have 60% chrome users that your demographics can be extrapolated from that figure. Make your own demographics surveys and ask these demographic questions manually when doing testing. This is the best way ot know if your assumptions appear to fit reality.

  3. Of course asking people to use software they're not comfortable with will affect usability. Whether it will "corrupt the findings of a test" is questionable; structure your test so that the browser is all but irrelevant.

    Unless your page requires use of browser-specific features like bookmarks I don't see why allowing your users to pick Chrome, Firefox or IE makes a significant difference as long as your site displays and performs comparably in all of them. Letting their use their typical browser is "normal" for them, forcing them to use otherwise will likely have a greater effect than the effects caused by different browsers regardless of user comfort.

    Consider A/B testing users with the different browsers if you feel this does have an effect. Browsers can certainly have an effect on the experience if you're asking them to do things like install add ons or download/install files. If it's a brief test you can easily run a within/between subjects design asking multiple or the same subjects to perform the same tasks in different browsers.

  • I don't think you should structure tests to ignore browser features, because in the real world people will be affected by these features and your website should take that into account. So structure your test to take into account that different browsers may affect your findings. For instance, test one group exclusively with one browser, and then test another with multiple browsers and determine whether there were any differences. Then you can move on from there to figure out how you feel about those results.
    – Rahul
    Commented Feb 5, 2012 at 0:35
  • 2
    a little note: If usability tests are drawn from what your target demographic, and if you instruct them to use the same browser they normally do, then statistically your results should line up with the real world. If they don't know which browser they use (cheep jibes aside, I do see this often) then you should probably go for IE (default), though I've often seen people use a browser that was installed by someone else.
    – Forthright
    Commented Feb 5, 2012 at 12:49
  • @Forthright They'll line up with the real world provided a stratified or sufficiently large, true random sample :). Hopefully they should be reasonably close to real world stats though, and certainly more representative than taking a population that uses various browsers and testing them all on one browser.
    – Zelda
    Commented Feb 5, 2012 at 17:50
  • What makes it more difficult is Chrome is preinstalled at Android and this means even not tech-savvy will use it (Actually my mother does. She has no idea what a browser is). And spending most time in internet? Mmmh, am I spending time if my Apps update continiously? Nowadays you are literally always online, not just minutes or hours. This sounds like a figure from old modem days.
    – FrankL
    Commented Feb 6, 2012 at 18:33
  • 1
    Android doesn't use Chrome (but it will soon)
    – Rahul
    Commented Feb 7, 2012 at 2:07

As some answers have shown, thats its very dependent and more likely a assumption, I would like to show a possible way how to characterize a user by more than one feature (browser).

You can look for installed

  • fonts
  • plugins
  • audio / video codecs
  • browser add-ons
  • google/yahoo toolbar (if detectable)

Additonal computer features like

  • language
  • operating system
  • browser
  • cookies on/off
  • timezone
  • time of use
  • screen size
  • viewport size
  • location

And all this data together, give you a very nice profile what sort of character your user is. You can test yourself here at Panopticlick.

Anyway, I think mining this data is out of our daily scope. But I think its very precise then, like who is browsing when, what, how. How customised is his/her computer system and in what direction if so.

  • An interesting thought.
    – PhillipW
    Commented Feb 7, 2012 at 10:38
  • A friend of mine told me, once some Google Advertisers asked for increasing their GoogleAds and he told them, they can restrict Ads for male, good income, interested in building loan contract, around 30 in a specific city. Isn't that spooky? That was before Google+.
    – FrankL
    Commented Feb 7, 2012 at 13:10
  • @FrankL: That was the case even in paper spam in the nineties. Google does have better (lower) false positive/false negative ratings, though. But I bet my sister-in-law (who does google powertools) will get those building loan contract ads, too.
    – MSalters
    Commented Feb 7, 2012 at 15:19

This is a great question. There will be a lot of inference made on my part in answering this, but I will provide my view and add references to the views of others where applicable.

1. Does age, gender, or profession affect a user’s choice of browser?

In short, yes. Let us first tackle age. In my opinion, the majority of users who are 50+ are more likely to use a Windows based machine, and with that, use Internet Explorer (the default browser). Younger users, who have grown up with computers, are more likely to experiment with software and find a browser that meets their personal needs and values.

I am going to tackle profession next. If you work in technology or media, you are much more likely to be aware that their is a choice of browsers, and as you are familiar with the industry and comfortable with technology, you will choose a browser which best meets your needs and values.

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.

Gender. I think it is safe to infer that their are more men in technology jobs than there are women. If you are in a technology role, as mentioned in profession, you are more likely to make a choice about which browser you wish to use. Therefore, I believe more men than women will make a browser choice.

Over half of the professionals in the United States are women. They account for nearly 50% more college degrees than men. With this kind of knowledge, how is it right that less than 20% of the working computer hardware engineers are men? Answer: it isn’t.

2. Are users of a certain browser more tech saavy and more easily muddle through
usability problems than users of the remaining browsers?

We covered “Are users of a certain browser more tech savvy” in profession during the first question. If you are more tech savvy you will be aware of browsers non-tech heads aren’t. Any browser that is not a default operating system install is more likely to be used by someone who is tech-savvy (unless it has been installed for them by someone who is technologically more advanced — but it could be inferred they will return to their ‘known’ browser choice).

The second part of your question “… more easily muddle through usability problems than users of the remaining browsers?”, it could be argued that tech savvy folk could be more forgiving of interface issues, and will be more likely to experiment and correct issues, gradually aligning their mental model of how a system works with that of the designers. However, I feel it is important to note that this is largely browser independent.

3. When conducting usability tests, will forcing the testee to use a browser 
that is not their preferred browser corrupt the findings of the test?

The chrome of the browser shouldn’t corrupt the test. There are a number of commonalities between browsers and standards which they should comply to. A test environment should use a browser ‘out-of-the-box’, that is, no extra toolbars, add-ins and panels. You could always put the browser in to full-screen mode to entirely remove the chrome if you have any concerns.

What you should consider is how the interface within the browser has been designed. For instance, if you have a progressively enhanced design which has very different views in, let’s say IE6 from the latest stable version of Chrome, if you have the time and budget, you may wish to consider testing across variants.

  • I just had a further thought. Microsoft have had to recently provide a screen with Internet Explorer to inform the user that other browsers are available tomshardware.com/news/Microsoft-browser-ballot-EU,9705.html — this may have migrated some users over to other browser vendors.
    – DigiKev
    Commented Feb 4, 2012 at 22:16
  • 4
    -1 Too many assumptions, and some bad advice at the end: putting the browser in full screen mode won't give you any information about how people actually use their browsers, so your test will be biased from the start. The question is about whether the choice of browser affects behaviour, so removing those differences won't fulfill the conditions of the test.
    – Rahul
    Commented Feb 4, 2012 at 23:24
  • @Rahul I don’t think the answer is not useful, and I disagree about your comment on the test being biased by putting the browser in full-screen mode. The point is you are testing the interface of the website or application within the browser, the test is not about the browser itself.
    – DigiKev
    Commented Feb 5, 2012 at 0:15
  • The web site/web app and the browser are intertwined. You don't have one without the other. So I'd agree that a full-screen mode is likely a bad idea for web site testing.
    – DA01
    Commented Feb 5, 2012 at 2:32
  • "Younger users, who have grown up with computers". Tsk ! They did have computers back in 1980... (the ZX 80 for one)
    – PhillipW
    Commented Feb 7, 2012 at 14:17

Just a quick little pennies worth here, this article can help in somewhat, especially the part that reminds you that Mozilla is built on the foundation of being free and independent. This would suggest that the demographics that fight for their rights and enjoy their freedom within the internet, would choose Mozilla.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.