>[mailto:thelist-admin at lists.evolt.org]On Behalf Of aardvark >> From: "Mark Cheng" <mark.cheng at ranger.com.au> >> >> Microsoft is a perfect example of how alienating users doesn't make a >> difference to the bottom line. > >no, but it does get them tangled up in all sorts of litigation and anti- >trust cases... > >and MS actually does a good job of accessibility in their apps, >when they do them right... Accessibility. Not useability or degradeability. Try running your Excel 98 Macros in Excel 2000. Or worse, vice versa. > >> My guess is have a larger buffer and not piss off 10% of their >> clients. Note that I said clients. If they piss of 10% of the people >> who hit their website well, so what? There are always going to be >> businesses that try to run on really thin margins, but if your clients >> are trying to do that on the web I suggest you get paid in advance. > >oh, i do that, but look at standard business models that aren't on >the web... alienating 10% of customers is not something they are >willing to do in the rest of the business world... grocery stores don't >do it, for example... competition is fierce... You used the magic word - customers. Stores/malls routinely adopt techniques to keep some users out. >is it 10% overall? 10% >per month? 10% of people who would have otherwise bought lots >of stuff? hard to tell... I don't know - Martin came up with the 10%. An illustration of the problem with statistics! > >> The basic problem with trying to predict web sales is that you can't. >> Caches, ISPs, hidden referrers etc etc corrupt the data that you can >> see, which, even worse, is stated in percentages for the most part. >> What is it a percentage of? US adult male users? Global users? > >exactly... which is why so much of this thread is academic... many >larger sites do focus groups and have qualified researchers tell >them what those numbers are, and how they break down... for >sites without those budgets, you have to rely on aggregate stats >and experience... > Actually, is it academic? aren't the aggregate stats saying you can reach 80% of a hell of a lot of people by targeting ie5+? That's a hell of a potential customer base as it is. >> What about the 5% of users that show up as unknown? > >those are all my friend Ed... it's what he does in his spare time... At last I know :) >> Businesses make numerous assumptions before they even get to the point >> of saying they need a web site. For a website which is driving sales, >> business doesn't care which browser is used to make the sale. They >> will care about ongoing maintenance, ease of updating, and the cost of >> putting it there in the first place. > >again, that's the rub... we were initially talking about sites that do >the transaction... it's been modified by the original poster to >become the marketing sites that fuel the decision to do make the >transaction... Yeah, but if I was advising a client on an e-commerce site I'd go for the highest level of encryption available in the browser population. For an ecom site security of comms between the browser and the server should dictate - business risk will dictate what level of encryption the client is happy to accept. >well, good research is better than assumptions... hell, my first >focus group taught me that... i know we have researchers on this >list, but you can get some statistically significant info on a site's >users in more ways than parsing logs... > You've heard of the phrase lies, damn lies and statistics? imho statistics, used properly and in context, can be valuable if you have a full understanding how the information was obtained in the first place. >> Sure, you can use server side languages to serve up the exact same >> page to each individual browser, just in case, but when you look at >> the effort of doing that compared to the additional percentage of >> sales you'd need to get from the gen 5, is it worthwhile? > >huh? you mean coding many pages or no? i don't code many >pages, so this is lost on me... > eg not serving images to Lynx, stuff like that. >> Choosing to use an old browser doesn't obligate a business to provide >> for that. Web designers need to tell / show the client what the site >> would look like in an uncatered for browser (or with JS off or images >> off or whatever). "Educating" clients that they need to support older >> browsers catering for a possibly small potential audience outside the >> businesses target demographic isn't good advice. > >no, educating the client consists of helping them determine the >stats of their audience, and showing how the site will and won't >function... i suspect you think i tell them to just support older >browsers, but i don't do that, i actually show them, and let them >decide... That comment wasn't aimed at you. it was a general response to a couple of earlier responses implying that web designers had a responsibility to encourage clients to design degradeable sites. I agree with your approach. This email may be confidential and contain commercially sensitive information. Only the intended recipient may access or use it. If you are not the intended recipient please delete this email and notify us promptly. We use virus scanning software but exclude all liability for viruses or similar in this email or any attachment.