Forum Moderators: open
Okay... that was rude. Sorry.
Yes, it does matter, because every time your site attracts 1 million visitors - whether that takes a day or a year - it means that 50,000 of them think it's naff. Every one of that 50,000 is potentially someone who will buy your product, click on an advert, or even just send a link to a friend.
Designing a site that looks good in all the major browsers is hardly harder than designing one that doesn't. Go on. Go the extra mile -- it's worth it!
(BTW: I have no idea how accurate the 50,000 figure is. The exact number isn't really the point of what I'm saying.)
So I don't think you should worry.
(As an aside, I wonder how many people's Internet experience is spoiled because of the number of sites that have not been written for them? They must see lots of problems.)
Server logs may not easily show that percentage, because a site that doesn't work on a minoity browser or OS just doesn't hold the traffic, so naturally you only see lots of page views for those you actually can serve.
And don't forget people who spoof their user agent as a matter of course, just to get any page at all page from the poorly coded websites out there.
I find it takes much more development time to accommodate IE4, 5 & 6 than it does to accommodate FireFox (growing like crazy) and Opera. But on most general interest sites I run, it's closer to 15% of the traffic that runs non-IE or non-Win.
And that's a number that is likely to grow this year. I don't lightly throw away 15% of my potential income (or even 5% for that matter).
I find the easiest approach is to develope a validated page template that looks right in a standards-based browser like FireFox. Then just a few tweaks for IE and Macintosh and it's all set.
But if I begin with a WYSIWYG approach for IE6, writing non-valid tag soup, then the repair work to acommodate all the other user agets can be overwhelming.
Here's my basic list of priorities:
1. Is it standards-compliant, or as close as possible?
2. Does it render well in IE 6? (still majority of browsers)
3. Does it render well in Firefox?
4. Does it render well in Safari?
5. IE for Mac?
6. Legacy (4x)?
I'd say the top 3 are critical, and it goes down in importance from there. There's always a cost-benefit analysis to be done with legacy browsers, and you don't want to seriously inconvenience 95% of the users for 5%.
For instance, say you're using best practices and standards-based code, with an eye for accessibility, and the resulting page in Netscape 4x doesn't render prettily. Your options might be:
1. Redo site in tables, adding 15k per page for all users.
2. Fork based on javascript browser sniff, adding 5k per page for all users (sniff code) and requiring two templates.
3. Leave it ugly in Netscape 4x
In this case I'd say leave it ugly in Netscape 4x.
The opposite extreme, of course, is a page that looks fine in IE, but due to their box model weirdness, looks like crap in other standards compliant browsers. In this case, I'd say do the fixes to make it look right across the spectrum, because the price of those fixes is significantly less and the benefit is significantly greater.
I've gone the other way around on a much simpler design and would estimate roughly 75%-100% more time to tweak for Firefox/Safari/Other from and IE driven code base.
It's actually more than 15%. Because you are also catching some of the customers that many your competitors are thowing away - they click back and move to the next site - hopefully yours.
Hmm, I've not run into many problems with sites looking crummy in Firefox. If it looks bad in Firefox, I'd guess that it probably doesn't validate, either.
Checking that Mozilla works is worthwhile. A November 2004 survey reported in eweek said Mozilla had 7.4% of the browser share. Firefox is the most popular browser based on it, with 5.6% of the total browser share. It's also gaining users rapidly, so you definitely don't want to cut those people out.
Netscape seems to be in decline (did they stop development on it?) so that may not be as important, although it seems to do pretty well with most of my standards-compliant pages. I no longer support Netscape 4.x because it's just too much work for too little benefit. At some point we've got to give the boot to really old browsers, right?
Apple's Safari only had .9% in that same survey, and Opera 1.3%. Small potatoes....
Of course, who knows if these stats are accurate?
IE: 90.28%
Firefox: 4.95%
Non-Firefox Mozilla: 2.64%
Other: 2.06% (with nearly a doubling of Safari stats because of a December security upgrade, something websidestory didn't think would be sustained)
Small potatoes....
If one of those "small potatoes" happens to be a webmaster who refuses to link to your site because they can't view it, the negative effect of that dismissive attitude could be bigger than you ever know.
The chances are that almost anyone would be able to view a standards-compliant site. Even Netscape 4 users can see the text of my sites. So it's not a matter of not seeing it at all. Or at least none of my sites have even had that problem.
The more important question we're dealing with is how many different browsers are you going to test with to be sure it looks great? In practice the differences are usually going to minimal.
If a webmaster couldn't view your site, at least they'd know what to do to see it. And we can't do anything about our viewers' personality problems!
I don't think we need to test every browser. So for my time and money, I'll test the top 4 or so, plus a few that might be important for special niches like the Mac OS9 users (IE5/mac).
Also, as others have pointed out, Firefox is also a growing market. Ignore it at your own risk.
I find the easiest approach is to develope a validated page template that looks right in a standards-based browser like FireFox. Then just a few tweaks for IE and Macintosh and it's all set.
That's the way I try to approach it. Since I've been using Firefox as my primary browser, I'm finding that it has made me more aware of the box model oddities of IE and quirks mode, that's for sure. <grrr> But it's still easier to tweak for those gotchas than to come at it from the IE end of things.
Gotta love the Web Developer Toolbar in Firefox.
Lisa
I bet most of them have Internet Explorer set as an alternative browser so that they can view non-compatible sites that interest them.
The fact remains that people who use minority browsers by choice are mostly savvy.
I'd have to disagree with that, at least in part. I had one friend (who uses FireFox) ask me if FireFox and IE "use the same Internet." Though he's highly intelligent in other areas of life, I wouldn't say that particular FireFox user is exactly "savvy" in tech matters! ;)
Apple's Safari only had .9% in that same survey, and Opera 1.3%. Small potatoes....
Small potatoes can add up to big chips.
Think about a market where 10 competitors are close to equal on most factors. You'd expect each to take 10% of the sales.
But what if only one of them catered for the 2.2% of small potatoes?
They'd each take 10% of (100 - 2.2) = 9.78% of the market.
Except for the guy with the free run at the 2.2% who'd take 9.78 + 2.2 = 11.98%
That's nearly 20% better sales than the other 9 competitors.
Factor in similar small potatoes (the site that works well for the 5% of disabled users; the one that downloads in time to catch the 6% of <56.6KB modem users etc) and you are talking a major competitive advantage for a site that works well for the small percentages.
This simply isn't a real issue.
The fact remains that people who use minority browsers by choice are mostly savvy.
That is true in my experience. And it's exactly why they have more clout than the bare percentages would make you expect.
So consider this for a minute, you go to your favorite fast food restaurant, and 1 out of every 7 times a customer comes in, the cashier yells "we don't serve _your_ kind" and kicks you out. After all, that's what you are telling your internet customers...
:-)
...it's exactly why they have more clout than the bare percentages would make you expect.
Excellent point, buckworks. It's that "Tipping Point" factor coming into play. A great book that I heard was one of the marketing bibles for Google during the early years, at least.
The fact remains that people who use minority browsers by choice are mostly savvy. They would have to be to consciously decide to use them. It follows that while surfing they will encounter many sites that are not displaying properly so they must be happy to live with this
I know that more than 95% of my visitors will be happy with what they see. Add to that the 2% or 3% who may see a minor problem but a usable site then I think I'm almost there using my existing methods.
I am a busy man and if I get the chance to adjust my site I may do so but it is not at the top of my priority list, nor should it be under these circumstances.
