20

I'm working on a brand new website and trying to figure out how to decide the oldest version of IE to support since I have absolutely no user information to look at. Most of the answers I come across for questions like this tell the person to look at their analytics. But in this case, there aren't any. Until the site is built and launched all I have to run on are general statistics about browser usage:

http://gs.statcounter.com/#browser_version_partially_combined-na-monthly-201404-201503

But this is from every demographic out there. In my particular case, my users have to make at least $50,000 a year. Unfortunately, Google doesn't help me with a search looking for a correlation between income and IE browser version. I suspect people who make about that much are on newer versions of IE. But I have absolutely no way of proving this until I have users.

So, how do I make decisions like this for users who don't even exist?

  • 7
    Don't over think it. Support the last couple versions, launch, and then see if there is a need to support older versions. Remember that 'support' is a fuzzy term as well. You can often support older browsers at least in terms of content delivery. – DA01 Apr 27 '15 at 15:55
  • 1
    Personally, I'd build for IE9+, tweak for IE8 by loading hacks with conditional comments, and only retrofit with even more hacks for IE6 & 7 if absolutely necessary. You'd almost never build for IE6 or 7, you just have to factor additional time for hacking it into shape if necessary. – Dom Apr 27 '15 at 17:28
  • 1
    Be aware of what your chosen framework supports. Ex: AngularJS 1.3 no longer supports IE8. – Moby Disk Apr 28 '15 at 4:09
10

Part of the answer may be dependent on market segment. If you're targeting Asia, you'll need to skew toward older versions of IE. Likewise for users that have overbearing IT departments that are behind the times.

As with everything else UX related, find users that match your segment and ask them about their technology. Way better than assuming.

Barring that, research your market segment. Is it JUST income that separates your users from the general market or are there other factors? If it skews younger or non-corporate, you're probably safe without a lot of backwards compatibility. If it's older and more corporate there's more of a chance you'll need to think about compatibility.

  • 2
    It's all about your target. For most consumer apps, you're safe to ignore the trouble makers (IE 6-8). MS has gotten aggressive with their legacy support policy: only the latest version. OTOH, if you're addressing a specific niche or a getting into a business context you might be dealing with some old school tech. As always, research your target. – plainclothes Apr 27 '15 at 15:53
19

I have dealt with this question many times, and it's all a question of cost effectiveness. You should ask yourself:

do the X% additional users that use IE 6-8 worth the Y% additional development cost?

In order to answer that question, you will also have to determine how costly it is to support every browser version, considering the fact that some features are easy to develop for old browsers and some are almost impossible.

These days I personally usually start developing for IE 9 / 10 + (depends on the complexity of the website), and add support for older browsers when I see large amounts of traffic coming from them.

I often refer to Can I Use to get browser usage stats and feature support.

  • I love that site! Also that really depends on the OP's immersive desire for old browser compatibility. The older you go, the less compatible certain code features potentially will be. – Majo0od Apr 27 '15 at 18:37
  • 3
    Generally a good answer, but the OP stated he doesn't have X, which can vary quite a lot between websites. – Prinsig Apr 28 '15 at 13:30
6

IE8 or not IE8

What I've found is that various shims and the usage of widespread web frameworks means that support for IE9 and above is fairly easy.

IE8, on the other hand is a completely different story, and many framework have or about to phase out support for it.

The usage statistics chart looks like this [source]:

A pie chart of browser usage, showing IE8 having 15.11%

And as you can see, IE8 and below are around 17% of the market.

Who are your users?

But if you look at the analytics of some services, you'd find that IE8 and below hardly scrap the 2%. This is because the service attracts users of that are fairly teach-savvy and work for themselves. For example, you can assume that users of Balsamiq will use IE8 less that users of Gmail.

Another trend is that users in big, bureaucratic, deep-hierarchy organisations tend to use old browsers more often. The usual suspects are (ones that I came across at least):

  • Big banks
  • Call centres
  • Charities

Many of these organisation got old machines, that are rarely updated, and are lock for any updates without expensive IT contracts.

  • The question mentions users who make at least $50,000 a year. For these specific users, worldwide stats like these are completely irrelevant and misleading. – benjaminjosephw Apr 29 '15 at 7:45
  • @benjaminjosephw, while I don't reject what seems to be your assumption, it'd be more constructive to provide a more relevant statistics or at least a stronger argument. Nevertheless, I wonder if in your view a world wide statistics showing 17% of IE8 usage, and one that would show 0.5% still will have no relevance? The statistics given is just to give a rough indication to the scale of IE8 usage. – Izhaki Apr 29 '15 at 8:46
  • I don't think it's just an assumption to say that a $50,000 salary indicates a specific audience subset which excludes the vast majority of the worldwide audience in your stats. I do, however, see your point that these stats can give a rough overview of general usage. Even so, I think it's better to make a well informed "assumption" than use the wrong data to back your case. – benjaminjosephw Apr 29 '15 at 9:05
4

When deciding what browsers not to support, understanding who's using the application, what the application is and where and when it will be used should provide a good starting point.

1. Who?

Understanding your demographic will give you some insight into what kinds of devices and browsers they may have access to. Income, age and profession will all give clues. It's difficult to find hard data to correlate these factors with device capabilities but even just taking the time to think this through provides a reasonable starting point.

2. What?

The purpose of the application you're building will suggest which contexts it's likely to be used in. Corporate contexts are imply access to different technologies than those that users might have access to in personal contexts. The type of application may also mean that inaccessibility for some is permissible or unacceptable.

3. Where and When?

Where and when users will be accessing your application is perhaps more crucial when considering device support as opposed to browser support but this could play as a factor. This would especially be the case when the application is intended for personal use in a context where you are confident the user has access to good browsers but they are also likely to use the application while at work or while travelling.

These questions will provide a reasonable starting point. Ultimately, you won't be able to judge exactly how the application will be used until you release it and analyse usage patterns. Building a good view of likely scenarios should help guide your initial assumption.

0

One factor to bear in mind (for private users) is the version slide: outside of business users, which may have old versions for in-house compatibility reasons, many (most?) people will have their software updated automatically, which, for the particular case of Windows/IE means: assume IE 8 for XP, 9 for Vista, 11 for Win7 and newer.

0

Up until a week ago I worked for an online education company whose entire business is based on offering online education. Students and faculty interacted online for assignments, discussion boards and quizzes. As a rule of thumb, any browser that had more than 2% of users was supported. Personally I would move that threshold to 10% and urge those users to update their browsers.

-1

My advice is to select your browsers and their earliest iteration for which you would expect functionality. Often this refers to IE6 being the test for all things to break.

Sadly I don't have any reference for this recommendation, other than experience. The major function fall off appears to rest with IE6. We never had any issues with Chrome or FF to compare.

  • While that was true long ago, there was a major functionality shift between IE8 and IE9 such that many won't support IE8 and under or charge more for it. – Rob Apr 27 '15 at 19:50
  • That sounds more like opinion, for which I've been marked down for above. I've never had huge issues with IE6+, as the answer/advice says, design as best you can and the rest will fall into place with a few hacks. Whether you want to foot that cost is another matter. – DarrylGodden Apr 28 '15 at 8:18
  • 1
    I would not say this is true any more. IE8 to IE9 was a huge upgrade, bringing support for things like media queries, SVG, canvas, multiple backgrounds, box-shadow, border-radius, etc. Not to mention HTML5. – superluminary Apr 28 '15 at 9:38
  • Trying to support IE8, retina displays and pinch to zoom devices with the same website involves significant time and effort. – superluminary Apr 28 '15 at 9:39
  • Well you design for your target not just for your stakeholders. – DarrylGodden Apr 28 '15 at 10:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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