Welcome to WebmasterWorld Guest from 107.23.37.199

Forum Moderators: open

Message Too Old, No Replies

Your Site Layout Might Be Killing Your User Engagement

     
5:55 pm on Oct 13, 2015 (gmt 0)

Administrator from GB 

WebmasterWorld Administrator engine is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month Best Post Of The Month

joined:May 9, 2000
posts:26187
votes: 970


There's a great piece published on one of WebmasterWorld's Administrators, rogerd's Neuromarketing site, on site layouts, and performance.

It discusses vertical vs. horizontal layouts [neurosciencemarketing.com], and the performance expectations. Heat maps prove how a trendy layout may just kill a site, even if it looks great. Not only that, but perception can make a difference. This piece is really worthwhile reading, imho.

In short, don't make it difficult for site visitors, but if you want to find out why, read the article.

It's that much tougher to win an audience, so once you've got them you want to maximize retention.

Have you used any of these ideas when you last designed, or are just about to redesign your site?
5:18 pm on Oct 16, 2015 (gmt 0)

Senior Member from FR 

WebmasterWorld Senior Member leosghost is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Feb 15, 2004
posts:7139
votes: 412


if you mean WebmasterWorld ? take a look at the CSS, and the source code, it uses more than one font..
5:26 pm on Oct 16, 2015 (gmt 0)

New User

joined:Oct 14, 2015
posts: 8
votes: 3


This thread is fascinating. I am firmly a desktop PC, multiple-monitor user. I have started to use my phone for composing things and getting minor internet work done by necessity only. But when I don't have a keyboard, mouse, and dual monitors, I feel like I'm handicapped. I only say this to show my bias.

I just assumed this "new design language" happened purely because we were moving to a tablet-based world (and I was just not part of that, so I had to suffer). I didn't realize *everyone* was suffering.

@lizardx, your numbered, point-by-point rant was so spot on. This has been infuriating me for a couple years now. It first really started affecting me when Google changed all their Play store pages to reflect this. Until then, I could just choose to not visit sites that had switched their layout to this trend.

The information density reduction in particular is incredibly annoying.
6:20 pm on Oct 16, 2015 (gmt 0)

Senior Member

WebmasterWorld Senior Member ergophobe is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 25, 2002
posts:8637
votes: 283


serif fonts, not sans serif


Tangor - I think this is partly age related. This is from memory, so I might be wrong, but I thought I read something where people over a certain age prefer serif. I think that age would be around 40 or maybe 50 by now. I'm on the "old" side of the divide and at least in print sans-serif just does not work for me. That said, it turned out that most people read faster in sans-serif, regardless of preference. Again, this is old and from memory. I think it was on the Usability News (from Kansas State?) years and years ago.
6:24 pm on Oct 16, 2015 (gmt 0)

Senior Member

WebmasterWorld Senior Member ergophobe is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 25, 2002
posts:8637
votes: 283


OK, scratch that comment. Alex Poole has put together a great review of evidence

[alexpoole.info...]
8:33 pm on Oct 16, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


AMC4x4, what struck me about the information density reduction on the new faddish sites was how incredibly bad it is for usability. But what really amazed me was bloomberg. If you are familiar with the bloomberg market, that is financial professionals who need to scan a lot of date quickly every day, from a variety of news sources. So for bloomberg to go from a one screen scan to get all the headlines (with an excellent mobile only version of the site) to a maybe, 4 or 5, maybe 8-10 screen scroll that doesn't even let you do a quick scan (keep in mind, the human mind only retains about 7 things at once easily from a page view) that one I simply could not understand, because it's impossible that was liked or found desirable by its users, who are all almost by definition power users.

I had not really dealt with mobilization of web sites until this year, because I thought it was a waste of time, but finally our user base was showing there was enough tablet/mobile use to justify it, but what struck me when I started using sites, and updating our code to be responsive, was what pure and utter bs this fad is really filled with. it's no problem to tap links, none at all, as long as they are far enough apart, and the font size is big enough. Now there is a danger with mobile users, as more and more sites go to low information density, massive touch surfaces, the normal training that would have happened to expectations and dexterity will not happen, and users will start to want dumbed down display surfaces. We haven't seen any sign of this yet, but our stuff is a work in progress, and we don't cater to younger users, who in theory would be MORE skilled at manipulating small things than older users, since they do it constantly, re mobiles, not less.

But I saw the same thing with netflix when they changed, say, you had a tab open with a tv series, it would show a normal text list of all the episodes, you know, like a table of contents, in, like, a book or magazine? Since those were text based links, they would, being html, show last visited status, which shows you which shows you have watched, for example. Then they changed to this aweful side scrolling list of big images, so say, you went from viewing 22 episodes for a year on one screen, including seeing which you had viewed, and under that, the next year, etc, to having maybe 4 big image boxes, which give almost no date to the user.

The comment that one should just say no to excessive js seems to be lost on many new web developers, and I'm starting to understand why, it becomes obvious when you start following web developer resources, almost everyone seems to be wanting to solve ALL problems with js, and I believe this is that age old web designer technical incompetence, where they don't know how to do server side programming. A few sites it does make sense on, like facebook, because really the entire page, almost, is dynamic, but they are the exception rather than the rule.

I see this more and more, where people submit forms using ajax simply because they don't even seem to know how to do it with html to a server side language, not because it's a good option. Same for writing out page content, that isn't dynamic.

ergophobe, despite lucy's protestations, the serif/sans-serif question I find to be pointless, users don't treat web pages like books, they if you are very lucky will read through one single long page, and scan a few others. So the real question there is: what scans better. What looks better? What I've come to realize that on web stuff, as with film, what is considered 'intuitive' is a complete fiction, these are conventions the human mind is simply trained to accept as 'natural'. I was fortunate enough to take some serious film classes where this was shown, how things we now assume are 'natural' and 'intuitive' actually came to be, like, exit left, enter right, for example,

Web stuff is the same, so any notion of natural and intuitive, simply means, conforms with expected patterns of use. Now, information density, that's an actual thing, ie, you either have it or you don't. This is where child sized fonts become dangerous however, because although people actually dislike that notion, they will use it, grudgingly, without knowing really what they don't like. That's a complicated issue.

But keeping it to just fonts, let's look:

I'm not concerned with variations on sans-serifs here by the way, the palette is bigger now than it used to be. As a base, I take the available palette to what windows xp shipped with, with font families that start, say, if you want to cater to apple users, the apple versions, then the windows, then the generics.

This is a fake test because I already by the way know the answer, but I'll do it anyway:

1. WebmasterWorld: sans-serif
2. bloomberg sans-serif, some headlines serif (I do this too by the way)
3. facebook - the number one site on the internet, sans-serif, of course, some headlines in serif.
4. ny times: they stick with serif, because we associate that with print media. ie, they want to look like print. this is a good decision.
5. LA times: same as ny times, they want to look like print
6. amazon: sans-serif, of course. I use amazon as a key utility, in other words, if I am curious how they are organizing their navigation, I see how they are doing it, though I have to note, with a major wince, their home page now is using hte big image chunky style too.
7. newegg - sans serif, of course
8. ebay.com - sans-serif, of course
9. google - sans-serif, of course

Since this covers probably about 80% or so of internet page views, it's a good sample. In other words, if you use the internet, this is what your brain sees.

But the real key here is this: ALL sites that want you to do something, engage, take actions, buy things, etc, in other words, the sites that interest most of us, use sans-serif. It's cleaner and easier to scan, less clutter. That's not an accident, it's why we use it too. Now, if your interest is in people losing themselves a touch in serifs, then by all means use it, but clearly the big sites who wants to make money from you doing something on their site are selecting what scans best, and leads to conversions, action, income. We've found the same exact thing, in fact, consistently when I tried to make our hx tags serif, I was after a while asked to switch them to sans-serif. Why? because sans was more legible.

One of the thing I no longer do, sadly, is pay any attention to alleged 'web experts' such as those studying the font style issue, why? because I have found that they generally simply do not have a clue, academics who are trying to show something or other, but who are always far far behind the times, and who often do not even realize these things are simply trained responses of the human brain, not actual rules of the brain.

But the great thing is, we all have direct access to web experts. In fact, nobody, as in our case, is actually more expert than us in our niche, in terms of conversions, seo, what our users like, etc, nobody, and everyone who has claimed to be were not, because the sad truth is, every demographic, and user base, has its own way of seeing and thinking, and it's up to us to FOLLOW them. User feedback, and particularly, users who fail to see or do something we think of as obvious (but it's obvious to us because we make the stuff, we can't duplicate the way a non tech user's brain works). So our feedback is: we love your site, great company, wonderful staff, blah blah blah. What we have not and will never see is users asking us this: please reduce the information density of your pages, we want only soundbites, and want to scroll a lot to access your information. In fact, we've been increasing the information density of the pages, not reducing it.

But to me, the true test was seeing what users of say, cyclingnews.com or chowhound did when faced with such redoes, which were clearly NEVER a/b tested, and I will bet you 10 to 1 that if a/b tests were done, the results proved that users hated it, but they went ahead anyway, totally ignoring the data. My strong feeling is that one driver of this is hierarchical corporate systems were managers are afraid of making the wrong choice, so they go with what others are doing, and that is literally the only thought process that happens at all.Then people start to create frameworks and libraries to meet this market need, and those frameworks and libraries are increasingly clunky, hard to modify defaults of (have you ever tried to debug minified js or css? that is NOT fun).

As an aside, to my eyes, print media that uses sans-serif looks really bad, it looks cheap, but screen media that uses serif is hard to read, and looks cluttered. That's because print and screen are two totally different ways of ingesting information. When you read print, you usually read a fair amount at once, when you read web, you often will read a page or less. So the way the brain works is pretty different, I can't say why serif appears to work better in print, it may because the extra curlies actually pull the letters off the white page, and make them stand out. I know that on a real book, printed very well, when I look very closely at the page, it appears that the letter actually is up from the page surface, but when you look at a computer screen, it's totally homogeneous as a surface, which of course is what makes in particular gray on off gray fonts so atrociously bad. LEDs just are not the same as physical ink on paper. I initially also tried to pretend they were on my first sites, but I no longer do that.

And of course, our friend physics helps explain this difference: a printed page is reflecting light, and that is what our eyes are designed for, but a screen is creating the light, which is a totally different way to see. And the reflected light is not built up from pixels, it's analogue. A good comparison is listening to a 64 bit per channel mp3 stream vs going to a live concert, that's about the difference I'd say in terms of the sensory input quality.

[edited by: lizardx at 9:14 pm (utc) on Oct 16, 2015]

9:13 pm on Oct 16, 2015 (gmt 0)

Senior Member from FR 

WebmasterWorld Senior Member leosghost is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Feb 15, 2004
posts:7139
votes: 412


(have you ever tried to debug minified js or css? that is NOT fun).

funny you should mention that, the reason I've been posting so far into my night / early morning this week is precisely because I've been debugging other people's* minified CSS and minified javascript..only way to do it is late at night, middle of the night, when no one is going to interrupt..files over 500kb and in one case 1.7 megs..<=yes..not typos..

Deep joy..might get it finished tonight..

This weekend, or early next week, it will be codeigniter, Jquery, php and database files..in Bootstrap..serves me right for not building it myself..had I done so it would have worked..

*I paid someone to do it ( I didn't have the time** ) what they made didn't work correctly ( yet the spec was crystal clear )..and now they have been paid they are AWOL..

**Fixing it is probably going to take more time than writing it from scratch ( without code igniter ) would have..
9:22 pm on Oct 16, 2015 (gmt 0)

Senior Member from US 

WebmasterWorld Senior Member ogletree is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 14, 2003
posts: 4319
votes: 42


How do you find web designers that know about such things. I have never had any luck finding a decent developer that is not backed up for months and charges a ton for a website. Very few designers do much more than put you in some theme they already have. Is it possible to get a simple 20-30 page non-ecommerce business website under $5000? Also have it done in less than a month. Some even take a month before they even get started.
9:27 pm on Oct 16, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


Hahhahahha. You can't spec this stuff out any longer unless you explicitly specify that they are not allowed to use crutch libraries. On a recent new feature we outsourced it, and I believe it cost 4x more for me to fix it and make it usable than we paid for the allegedly final product. So the only protection is that payment depends on no use of such crutches. I've found jquery itself, sans the endless bloated plugins to do simple things that take literally 40 lines max of code to do in straight jquery, is pretty good, that is I would say an actual improvement on a technical level. All I need to see on bootstrap is github, which sponsors it. Chunky, blocky, it's not a bootstrap, platform css was a real bootstrap, all it did was create some basic things, but even those I've found largely useless beyond a few utility classes. bootstrap is the worst garbage I've touched personally, because it makes all these design decisions for you, and a lot of them are really bad. But as noted, that's github, who are the poster child for childish bad pages and style.

As the brain does with pain, I'd actually forgotten it was bootstrap I had to fix, and when I saw, in the end, after dumping ALL the bootstrap code except for what we used, which was I think about, as usual, 40 or 50 lines of CSS, I realized that this garbage is not for making websites, it's for developers who can't be bothered to do their jobs right.

What I see is very simple: developers are increasingly outsourced, outsourced work wants to do the job quickly, and move on to the next. So they deliver using these premade utilities, and let me tell you, when say, jquery drops a core function, say, between 1.6 and 1.8, which they did in our case, and you install the current jquery 1.x latest, poof, there goes your shiny new application, and you have to fix it yourself. And jquery I view as one of the more mature and adult things, but they also drop functions, and not minor ones, and force you to reprogram them manually. Just imagine what happens when google gets bored doing react.js because angular has taken over the market, and you've committed to react. I expect that to happen within a few years, by the way, due to the winner take all faddish nature of these frameworks etc.

Again, what astounds me is how major web sites, not small players, use 3rd party hosted libraries, if you surf with noscript as I do, you see this all the time, you enable the site, and most of it still does not work, then you have to add more and more to just get a form to complete, from 3rd party vendors. They can't even be bothered to actually host the core site files themselves. This to me reeks of outsourcing, and I believe it's a big part of why in particular large corporate owned or controlled sites are going this route, it has NOTHING to do with the end users, and everything to do with short term corporate hacks and low skill web developers trying to make some fast cash.

My guess is, and it's a guess since I don't closely follow this stuff anymore, is that as the web consolidates and becomes more corporate, chowhound is a perfect example, the sties themselves stop mattering. Well, ok, it's not a guess, I have a friend who works at a big corporate website holding company, and you can tell they have no real passion or care for the sites they bought, that were almost always made by people with passion for the topic, then decided one day to cash out, and were bought out by these groups, who lack the passion, and can't grasp it was the passion that made the site good, not the userbase, the user base being simply the response to the passion. But corporations can't really go that route, because you can't buy passion in most cases. so they outsource, make their sites look like the other fads, drop the quality, etc.
9:34 pm on Oct 16, 2015 (gmt 0)

New User

joined:Oct 14, 2015
posts: 8
votes: 3


@lizardx - OMG I had totally forgotten how much I hated the new Netflix layout. Also, interesting that I also have a film background (majored in it oh so many moons ago).

I recently had a discussion with one of our lead developers (we're a storage migration company, and this particular developer is responsible for our GUI) about how I hate the whole context-based GUI fad that's happening now, where functions in those "big buttons" are only available when they're necessary, and are hidden when they're not relevant. Coming from a classic menu-based age, I find I want to know where EVERYTHING is, whether or not it's "turned on" so that I can find it later if and when I decide to use a particular feature. In short, I want to see EVERYTHING (within reason) and then decide for myself where I want to go. But it seems with this new paradigm in software anyway, the trend is to only show what's necessary. I think this was initially something that had good intent, an attempt to unclutter screens for people who used smaller touchscreens (tablets and phones), but the simple fact is that in order to design EFFECTIVE systems like this, you need to know HOW people use the software/GUI. Your quote here is fitting to this: "My strong feeling is that one driver of this is hierarchical corporate systems were managers are afraid of making the wrong choice, so they go with what others are doing, and that is literally the only thought process that happens at all."

I think what we came down to was that when designing context-based GUIs, there has to be a psychological component to it and a true deep-dive into HOW the application will be used, and some real thought put into this, and trial. I don't think a lot of designers go that far. So perhaps it's not the concept I despise so much as its generally horrible execution at the hands of those who can't be bothered.

As this applies to web design, my issue with Netflix and Google Play App Store (the web-based one anyway) is that they have REMOVED data that was instantly available at-a-glance, and instead put it behind a touch/hover, if they even included it at all. This seems to be happening more and more.

Please. I want to SEE MORE than that at a glance. Why is it that these app/web developers think users can't handle that much data at one time? If a design is cluttered, perhaps some extra time is just needed to figure out a way to represent that data better? As you mentioned, I don't think anyone ever asked for lower density pages. That smacks of upper management mandating something because they heard it from some other outlet or because everyone else is doing it.

But what do we know?
9:54 pm on Oct 16, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


How do you find web designers that know about such things.


Well, the first you actually have answered here: you don't use web designers. Web developers use web designers to design the look and feel of the site, ideally with as little code as possible.

Then the developer makes it.

In my opinion, it's quite simple. You have two types of people on the internet, maybe 3, one, they will fail. That's most people. They are highly concerned with the costs for the website, because they don't have a business model, plan, or budget, and they also make decisions that make them fail.

So they don't matter, there are endless wordpress blog mutated to cms designers who will happily create a fine lego block site for you, for a very low price. But even that, sadly, isn't the fact anymore, a happily now ex client of mine went that route, they got your junk wordpress site, with all the extensions plugins and addons, and it cost them about 10k, and that's for a pure cr@p site that just looks kind of flashy. I laughed, and praised the gods that they were no longer my client. That client has always had failure on the web as their core business model, they excel at failure, I know because I was with them a long time, and they had an instinct for it. It was one of the more joyful moments of my developer life when they recently showed it off to me, and noted, oh, it's having this problem, this display issue, etc, and they can't get it fixed, and i said, shortly and sweetly: I do not touch or work with wordpress sites, for any amount of money. (pro cycling shops, as an incentive to hire good mechanics, will post on the ad for mechanics: no cheap bikes accepted for repairs. That's because working on cheap bikes is hell)

But your question to me says it all, anyone who is not crystal clear on the difference between a web designer and a web developer is going to get shafted, low bid, junk.

Developers use designers. A designer is, in a real site, less than 1% of the total job. On a smaller site, they might be 5%.

My feeling after having done this for too long is that the notion of treating a website as a one time fixed expense is the recipe for web failure, so anyone who does that is already determining that they will in most cases fail.

On the other hand, if the site owner realizes that a website is a long term business investment, just like office rent, salaries, and other expenses, and uses that website as part of their business, well, then, your question is largely irrelevant, because no web presence on the internet ever had that issue, ie, they spend and spent money on their web presence, and do so continuously. There's exceptions to this, like small time vendors, people who run nice small businesses, who just pop in a quick templated site that uses hoster run shopping cart pages, they are low budget and generally will just find their way to their solutions, if they are going to succeed.

But the notion of whipping out sites cheap, that's a strategy for failure in my opinion, sure, you can find people to do that for you, and people who want to go that route should, it's a given they'll never compete with anyone serious after all, so who cares what they do?

Very few designers do much more than put you in some theme they already have. Is it possible to get a simple 20-30 page non-ecommerce business website under $5000?


To repeat, you MUST learn that anyone who calls themselves a web designer yet feels they have the skills to create websites that are actually worth having and operating is confused. It's fine for them to be confused about this, but it's not fine for you to be confused about it if you are the one hiring talent. If you met me at a party and asked me if I was a web designer when I said I worked with web stuff, which is what I usually say to avoid such discussions, I'd get a look of pure and utter disgust on my face, and walk away, and if they were threatening to try to get me to make their personal site for them, I'd add: You can't afford me, then I'd walk away.

I have tremendous respect for high end, good, web designers, who understand what being a web designer is (it's not really hard to grasp, the term describes it, it's someone who designs the pages). I can't do good design, my stuff all looks like it was made by a web developer, like here, WebmasterWorld does for example. Well, ok, it looks better than that, but it's similarly functional.

A designer designs what you tell them to do re the requirements, you look at it, and stop when it's got a good look. The developer, programmer, etc, builds it, after thanking the designer and paying them. Some decent designers actually can generate nice css designs, and that's wonderful, though you generally have to spend a lot of time fixing the css, and de-genericizing it.

Leave the brochure sites to the beginners who will evolve in two ways usually: they will fail, or they will come to understand that you can't succeed in a business that is highly competitive by not competing, ie, by not taking your projects seriously and as real expenses crucial to your business. With the exception of the many nice small honest online merchants who have simple requirements and tiny budgets, they generally don't need a lot and can get by with simple cart sites.

I used to show one client certain pages when they would ask, well, why can't we look like x site? the site was the old ibm.com home page in that case, which was a fairly famous page, they had spent about 250k on the design and execution. Which is what I told the client, lol. That shut them up a bit. So I delivered a similarly functional but not as good looking version for a few k.

5k for a 20 or 30 page site, that's easy. Good cms, that you understand, and can maintain, and that can be upgraded, with clean templating and snippets/chunks. Easy. That's assuming the client has their stuff together, and doesn't fall into the 'oh, you mean, I have to actually create the content? I thought you were going to do that!" which often stops the project dead for many many months. Actually putting up a site though, about 1 month is right. Assuming a custom template, which is my preference because it's easier to work with, a clear mobile responsive strategy, and a fairly clean structure, and no complex backend requirements, why would anyone need more than 5k? If you did it routinely, you'd have most of the tools already there, and just would modify it, so yes, 5k is about right. If you are more the core IT web guy, and don't do it routinely, like me, it takes longer and the job is more customized. But it's still about 1 month to make a site if the design and content are there. Less if it's simple, more if it's got features that are complex.

The reason the good guys are booked up is because they are booked up, that's not rocket science. Good car mechanics, dentists, etc, are also booked up. So are good contractors. Welcome to reality. That's because they are good. But I've NEVER seen a client who was actually ready to go re content in a month, not even remotely close to a month.
10:21 pm on Oct 16, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


Re web developers etc, you can create a short list of questions, and use it to get rid of the bad ones:

1. do you believe that a blog platform named wordpress is a cms?

2. do you know what a cms is?

That gets you past about 95% of people at this point.

If they pass these questions, you could ask them about what cms they use, and why. And when if ever they would use a js framework, and why.

For example, if they use drupal, which is a great cms for massive sites like whitehouse.gov, that have constantly updating pages, etc, ie, the scrolling news type of site, but upgrades horribly, and is hard to work on, then you could ask why they do that.

Then you could ask them about their templates, and how they do it. And their css, etc. And how much js they use, and why.

Basically anyone who can rationally tell you why they do x instead of y is a likely candidate.

I find that learning cms's is VERY HARD, so I picked the ones that I found were best for me as a developer, never regretted that decision by the way. They all have downsides, my preference is a pain to update, but it doesn't fail, unlike drupal and others. But wordpress is an automatic exclusion, instantly, because wordpress makes a fine, albeit cludgy and slow, blog platform, IF you use a true custom template that is well done (never needed to update mine in 10 years after doing it), and IF you do not use themes and other such stuff, which will always fail on an update. But wordpress 'pages' were never more than a secondary addon, which has been so abused and distorted that people stopped realizing that the pages were just some nice static parts of your blog, not a cms.

So you can just walk away smiling if a web person tells you they make sites in wordpress. Most of our vendors for one of our products, by the way, use wordpress with the same exact db plugins and output, it's comical, you'd be amazed at what runs under the covers of some of these very well to do companies, lol.

As an aside, we for one product used 5 vendors, all of which were totally and utter amateur, then finally one got pro, and spent some money making their tools, and presto, they are now the industry leader in their niche. No surprise there.
12:19 am on Oct 17, 2015 (gmt 0)

Senior Member

WebmasterWorld Senior Member ergophobe is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 25, 2002
posts:8637
votes: 283


Many debugging tools for JS and CSS will deminify now won't they? I believe both the Firefox and Chrome tools show the deminified source.

If it's a library, though, it usually makes life much easier to switch to the dev version with full source.
2:04 am on Oct 17, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


just gz the non minified files, same difference. As soon as you do that, google page speed tools stop complaining about them being non minified. This is another example of the encroaching incompetence, I think, it's because it's just too hard to compress your text files properly via the server for most users. For super heavy volume sites, sure, that few kb saved might make a difference, as will not compressing all those requests, but how many sites are in that category? 100? maybe?

I can only do the most rudimentary css/js debugging in firefox, for actual work I need the file itself. But by sticking to only one, core jquery, I solve the problem, and move on, there are better ways to spend time. Remember, one single image that is surplus, aka, every single box image on these new flowing box layouts usually, is bigger than any amount of kb you save minifying code.

One of the few advantages with getting older is not falling for silly things that don't help the end user or our site or me, the developer, and focusing on the things that do.

apache, 2 lines of code, and all your text files become tiny. Or is it 3? I forgot. So many new fads are just silly I find nowadays, I really thought the web would blossom with skilled users and developers when html5, css3, jquery 1.x and 2x became prevalent, but that's just not happening, it's VERY rare nowadays I see a site that actually makes me want to check out how it was made, usually the only thing that makes me curious now on site redoes is what they used, so I can avoid it. It's kind of sad, maybe that too long delay for MSIE to be usable for html5/css3 was just too long, I don't know, but something very fundamental changed in web dev since I spent a lot of time at WebmasterWorld in the past, and I honestly did not expect to see such a radical decline in skills and excessive use of crutch libraries as a poor replacement, this really caught me offguard, just goes to show, it's hard to predict the future in some areas.

On the bright side, nobody really talks about xhtml anymore, and xhtml 2 is dead, so that entire soon to be legacy branch of the web will fade away, where it belongs, thank heavens for html5 and css3 anyway.

Looking through a few thousand lines of code to find the 20 your outsourced dev actually used, now... that's not fun, but luckily it only takes someone paying me to do that a few times before they realize it wasn't a very good way to save money, lol.

But this is all connected, I think the boxy layouts, cludgy chunky designs, and premade minified js and css libraries, are all directly related, I don't think it's a coincidence at all, you can see it, easily, all the methods they use are the same, and it's all about generic fast dev time, not quality.

I started doing this to do quality stuff, and that's how I'll end it too, either that will have value or it won't, makes no real difference to me, this work is boring, so I'll do it until I don't.
2:46 am on Oct 17, 2015 (gmt 0)

Senior Member from US 

WebmasterWorld Senior Member tangor is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Nov 29, 2005
posts:9927
votes: 974


Gollies. Take a nap and come back... the kiddies have been busy today! Fun stuff, everyone.

Corporations end up too far removed from the dev team, but they are the one who pay the checks and answer to stockholders, thus can make odd, or even idiotic decisions. The only way I work with those types is as sub contractor for a SPECIFIC then take the money and run ... I know they will screw with perfectly good code and create a mess.

This Fonts conversation is interesting, but truly has no value in that serif and sans serif are both perfectly readable and usable. HOWEVER, there are perceptions in the audience between both, and that should have some consideration. My ecommerce sites are sans serif. My tech or informational sites use both, with serif for the general body text and san serif for either headline or blockquote, and monospace for specific code, or tabular NOT in a table data. News/entertainment/travel sites get serif as a nod and wink to "brochures" and "class", but the reality is that all sites around the world might have better legibility if the much maligned (with unwarranted prejudice) Comic Sans was used. That's almost humor because Comic Sans really can be quite useful (I use it for image captions in the 9-12pt range)

These font statements are "designer" attributes. Developers are the meat and potatoes part and sadly there's been entirely too much green peas inserted (JS) in recent years.

A website that wishes to be interactive needs something more than just html and css. Me, still Perl, the swiss army knife of the web, though php has got in the way simply by uptake. These are valuable and necessary, but dang it, JS is not (if you have either of the two mentioned). And UTTER reliance on JS will make some sites go dark to the visitor who is using a script blocker as MANY of us, even those who love webmasterworld, routinely do.

So, in the world of site design, layout, etc., the web dev needs to consider ALL aspects of the user as well as the content and the job at hand. These days of script attacks, malware by ads, etc, security conscious surfers are wearing not just Belts And Suspenders, they have Condoms On, too. Uptake on these user side tools (script killers, third party denials, ad blocking) are about 25% of the USA market and approaching 29% in the EU market. So, if your layout relies on JS to get 'er done, your site will not display as you intend. Just thoughts to keep in mind.
9:33 am on Oct 17, 2015 (gmt 0)

Senior Member from GB 

WebmasterWorld Senior Member 5+ Year Member Top Contributors Of The Month

joined:Sept 16, 2009
posts:1085
votes: 80


@lizardx - thanks for your 'rules' post, made me laugh a lot.

I think this is what happens when you go 'mobile first' and use a framework. I'm still learning about responsive but the thing that's struck me is that the relationship between font size and, for example, margin / padding size should not be a constant. The relationship even between <h1> <h2> <h3> <p> should not be a constant. So as you scale up and down you need to look at the the proportions and relationships between all those things. And it's not just multiple screen resolutions and height/width combos, but multiple browsers on multiple devices rendering even standard fonts, or the application of <strong>, with noticeable differences.
3:40 pm on Oct 17, 2015 (gmt 0)

Senior Member from US 

WebmasterWorld Senior Member ogletree is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 14, 2003
posts: 4319
votes: 42


@lizardx I guess I was being lazy when I wrote that question. I am very familiar with what you are talking about. I have worked in digital marketing since 2003. I have been posting and reading WebmasterWorld for over 10 years. I have even been a speaker at Pubcon. I have spent a lot of time over the years managing developers at other companies and have even trained many of them on how to code for SEO.

I am now on my own and don't have access to staff designers and developers. Part of the problem is finding somebody that is good without spending a fortune. I know they exist it is just a lot of work to find them. It truly is a needle in a haystack. Even if I were to just look for people that charge $10,000 plus for a website most of them suck. Having lots of customers and charging a lot does not get good work. Also it is rare that a company will even let you talk to designers and developers until you sign a contract. They want you to speak to the sales people.

I guess I could just find a designer first and then find a developer if I want to work with individuals. I guess I have made the mistake in the past letting one person do both.
6:07 pm on Oct 18, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


ogletree, it was more for people reading this thread than you, though I was surprised to see you use the term 'web designer' to indicate a web developer. It is however, as a person working with the web, to be clear on what one is doing, and why. Basically, we have two options: one, you make something as a one time thing for a client, and have no real relationship with that client. The other, you are creating a business tool for a business, and you are involved in that business. This is also known as a successful web presence, and company. Involving yourself with people interested in succeeding on a serious level is much more interesting and long term profitable than doing one off jobs for people who will never succeed, I've tried both ways, lol, and have no interest in doing the latter any more.

Almost all companies that start on the web, or were early on the web, and are successful, run real websites. I call a real website something that exists as part of the business, or the core of the business. The business of churning out short term projects for clients who will be gone sooner than later just isn't worth doing, though people will do it, and charge whatever they can get in the process. I'd rather work for one serious person than 20 flakes who will never succeed on the internet, if only because it's so darned boring doing the same thing over and over and over again, for no real benefit to anyone beyond getting some dollars.

Tangor, good comments.

In a sense my hope is to plant even the smallest seed to make people question the current fad of bad usability web design and development.

Re the notion of mobile first, that's what astounded me, I'd always assumed there was some validity to make sites bad, huge fonted, etc, but when I finally broke down and got a cheap mobile device, and started working on mobile code, I found immediately that me, with my bad eyes, had no need at all for these mega sized fonts, so at that point, it started becoming clear that it was very likely the package of libraries and frameworks defaults that were created the look and feel, NOT designers, and NOT actual decisions made by competent developers.

I guess I could just find a designer first and then find a developer if I want to work with individuals. I guess I have made the mistake in the past letting one person do both.


This is absolutely a mistake, the two skill sets are completely incompatible, they always have been, and they always will be. One is right brain, the other left brain, though both have to understand each other, and the basics of css driven design. But honestly, the designer really does not need to know anything about code, all they need to know is how to make a clean professional look and feel photoshop template, then slice out the image components, which should be few and far between, then present the developer with that finished product. that's my preferred method, though I have gotten reasonably good css, but that required a serious redo to make it responsive a few years later, so I tend to think now that the responsive component can't be tacked on.

The formula 'mobile first' depends on your user base, google developer says that, as if it's some rule that exists out there, but it's not true, it should, as always, be based on your user base, and what they do. but mobile is easier if designed in from the start.

I'd find the developer first then work with various designers if I were in your position, the developer makes the site, the designer makes the site look nice, so the specs should come from the site requirements and the developer.

However, re the real topic of the thread, it's important to realize that your site layout, and information architecture, and other key decisions, may truly end up killing your site.

Here's two sample letters to the editor responding to the chowhound story:

Flickr got Screwed First
it's worth noting that by chowhound general manager Georges Haddad's own account, he was also responsible for 2013 Flickr redesign, which Flickr users hated. The redesign prioritized slick looks over usability, undermined the community aspect of the site, increased loading times, and generally ignored things people most used and valued about flickr

That person nails it. Users hate the stuff, they may grudgingly still access the site, but the engagement vanishes. Think windows 8.0, Vista, etc, for good examples of this type of very bad corporate decision making.

I liked this letter because he nails almost every decline in quality, though he doesn't know why, the slow loading of course is because the js that creates the pages, the excessive dynamic parts, probably one of the major frameworks, all the pieces pulled from all over the web, simply take a long time to load, first, then build the page, churning those cpus and draining those batteries in the process.

Goodbye, Chowhound
I am a long time user - back when we would send money to Jim Leff to keep the site going. When I tried the beta version of the redesign, I was highly underwhelmed. I tried it a few more times to get dining info for an upcoming trip. I am dropping out of the Chowhound community as it is now.


These types of comments are exact duplicates ot the cyclingnews.com redo, which I took a closer technical look at when it happened because, as the first letter says clearly, there was NO benefit to the end users, everything was worse, there was no improvement in any area. I know speaking for myself, the engagement I had with that site, which usually was 5 to 10 articles read a day, went down to a few articles a week. For an ad driven site, that's a disaster, but I think that's what happens when a site that was created by and for passionate fans etc, or whatever the site topic, goes corporate, it loses pretty much all of its ability to really perform for its user base.

If confused on this, no need to look far, just take a look at myspace.com, which was bought for about 1/2 a billion by a clueless media corporation, and was sold for a few million (35 million I want to say) to a group that hoped to find some way to monetize the name and pages. I believe that's one of the better known blowups, and that was almost certainly caused by the corporation that bought it having no clue how to actually engage the users that created the site.

My feeling is that users hate this stuff, but can't really verbalize that hatred, and the site may simply have no real competitors, like bloomberg, so they just bite the bullet and accept the degraded functionality, but don't like it. But for people wanting to succeed now, I have to suspect there is a big base of people who want to be treated like intelligent human beings, and will reward site owners who want to succeed for that. Then there are others, like uber, who are really mobile only, with a site that does load on a desktop, but that's about it.

As I noted, I know a guy who works for one of these big web site holding companies, and the total lack of passion and interest in the actual web presences is almost astounding. There's a reason sites like google, facebook, etc, succeed long term, and others fade away into obscurity. Once you start this downhill slide, you grow unable to attract good talent, everyone can sense the decline and death spiral. Who wants that, except for a paycheck>
6:34 pm on Oct 18, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


One example. We recruited from two pools of users and segmented
- Pool 1 - random users recruited by a usability service
- Pool 2 - actual customers

Results? Markedly different click patterns, heat maps, scroll maps. Not surprisingly, the typical users recruited randomly to take tests on a usability site think and act a lot like me. Customers... less so.


this is ergophobe's early comment, and I want to suggest that you give this some serious thought, we'd been using those types of services too, and were about to do some changes to our core tools based on their feedback, but I had always felt that was a fundamentally flawed idea, and this comment shows me why.

One thing I'd correct, I don't believe usability services use 'random users', I believe they use people who do this for money, ie, totally non random users. But even assuming they are actually not people who make money testing sites, what we've found is that no generic formula works on our users, who reflect themselves. This was the most surprising thing I have discovered over the last few years.

I don't believe there is such a thing as a random user. I don't thing the 'average' user of newegg is the same as the average user of amazon, for example, and only amazon can really know what its average users do and how they react. Learning this is a time consuming process, and not particularly easy, unless of course you do silly things like listen to your users, adapt to their feedback, and watch their behaviors (users I have found have a VERY poor ability to explain why they react or act a certain way on a page, but tracking their behavior is a good way to see for yourself what works and what does not).

Also kudos to the first long comment by leosghost, which contains I think far more truths than a certain part of the web consultant industry wants to admit, which is they don't have a clue what they are talking about. That's been what's I see every time we've hired one of them. All they know is crude generic formulas, and after doing the real work ourselves for a few years now, the results are incredible, and TOTALLY non reproducible on a generic level. Conversion costs, down about 4 times. Traffic, up almost 100%, conversions, up 500 to 1000% (and this during a very bad time for the industry the site is in). Core marketing tools that were promising huge growth and conversion of borderline users, almost a total wash. Consultants for these services, virtually worthless, if not counter productive. Claims for layout ideals, totally wrong, negative, and counter productive. If I'd kept a list of all the statements made vs reality, it would be pretty impressive.

But the ergophobe observations, that's an area I had not considered.

The one positive I did experience was a young woman outsourcer who is the best bug and glitch finder I've ever seen in my life.
3:59 pm on Oct 19, 2015 (gmt 0)

Senior Member

WebmasterWorld Senior Member ergophobe is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 25, 2002
posts:8637
votes: 283


One thing I'd correct, I don't believe usability services use 'random users', I believe they use people who do this for money, ie, totally non random users.


Yes, I thought of elaborating on that but was trying to be concise. You got the essential point though: these users most definitely have a strong bias and you must be careful what conclusions you make from them.

I think they are good for proving somethign is a fail, but not necessarily a win. In other words, if they can't use your site, that's a bad sign. But there are exceptions - a B2B site that makes a small number of high-value sales to industry insiders may fail with these users and win with the industry insiders who have many $$$$ to spend on this purchase.

It is however, as a person working with the web, to be clear on what one is doing, and why.


I have so often had "the conversation."

Caller: "Hey, I'm looking for a web designer and X gave me your number."

Me: "Let's talk about what your needs are. I work with designers, but I'm not a designer. I'm more the person who builds the website based on the pretty pictures a designer produces. I can build a site for you myself, but if you want a custom design that will look nice, we should budget for a designer. They have all sorts of skills that I don't have and never will have. Design is really a profession until itself. Good designers are quite expensive and most that are good won't build a site of any complexity to completion. I don't have their skills, but few of them have my skills in much the same way that few interior decorators have solid plumbing skills and few plumbers have solid interior decorating skills."

And what exactly is web design? Because my analogy is flawed. It's more than interior decorating, but less than architecture. I think in the web world, the designer and developer both wear architect hats (ideally), but split the role of full-service architect, depending on the skills of each.
2:34 pm on Oct 20, 2015 (gmt 0)

Preferred Member from BG 

Top Contributors Of The Month

joined:Aug 11, 2014
posts:547
votes: 174


I understand why you guys as devs are so upset by wordpress and Joomla and all the rest of the BS "CMS" platforms out there. However realise that for a SMB more than 1000$ for a website is considered big investment. Now I can talk to them about laravel all day long but what they actually need is 10 page website that they as sub average user can edit, I'd just install a good stable Wordpress theme, or re-use one of the custom templates that me and my team created and know that they are secure and just work on the front end to get the visuals.

Supply and Demand folks Supply and Demand
3:19 pm on Oct 20, 2015 (gmt 0)

Senior Member

WebmasterWorld Senior Member ergophobe is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 25, 2002
posts:8637
votes: 283


I understand why you guys as devs are so upset by wordpress


Actually, that was one person above who said that. I personally try to push people to the minimum viable solution and that often is Wordpress with an off-the-shelf theme. Many people have ridiculous assumptions about how they just need to put a site online and the online business faucet will open. The problem I see more often is people with a completely untested business idea that has never brought in a dollar and they want to hire a professional designer for a completely custom site because their product is so awesome that anything less than an amazing site will not allow them to showcase that awesomeness. I find myself asking them "So... you've never actually sold one of these, but you have a $7000 budget for a website? Is that really how you want to spend $7000 of borrowed money?"

I used to build my own CMS, but now think everyone should be leveraging the work of others to provide the best solution. Sometimes that's a CMS. Sometimes that's a dev framework like Symfony or something.

I used to be hung up on light code that validated and all that. The thing is, most people want to be up and running and that *is* the right focus. Once the cash stream proves itself, you can go in and clean up the theme on most CMS and spend all the money on something custom.

So I think you should launch cheap and spend the money when you can demonstrate an ROI.

That said, I think where lizardx is coming from is that a lot of nice looking Wordpress themes can be next to impossible to work with, sort of like the cheap bikes example he gave. There's nothing inherently wrong with Wordpress (obviously, it powers some of the largest sites on the web), but there are things inherently wrong with a lot of off-the-shelf Wordpress themes and plugins that can lead people to disaster. A friend just took over managing someone's Wordpress site and it had 131 plugins. Whoever built that should be sued for malpractice!

But back to the original point of the thread - it's a matter of priorities. Again, if you have no proven business or your business is not primarily online, then the stuff we're talking about is optimizing around the edges. If you have a medium sized business, though, bumping up conversion rate by 0.10% could be a million dollars a year and then minor glitches in the user experience are worth spending quite a bit of time and money to fix. If you're Amazon, I can't even imagine how much 0.10% increase in conversion rate is worth. A billion dollars?
7:41 pm on Oct 20, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


Nutterum, I thought I was fairly clear, people who are not seriously interested in running a business where their website is a core part of their business should use whatever they want. Now, saying that, I also have to note that ALL users I've seen get shoved onto low cost budget options, without exception, ended up with absolutely nothing to show for it a few years later, ie, it was a total loss.

ergophobe has it fairly right, try to determine who the person is before starting. People who will never be using their web site as a significant web presence or income source should allocate however much money that low engagement level is worth to them.

I don't work with people like that anymore if I can help it because it's frustrating and boring and they are better off in the wordpress churn market, though the costs there can surprise you, I know I was surprised to hear what my ex client was charge for, but they are the worst clients in the world, and I'm sure their non stop unreasonable demands and nitpicking drove the price up rapidly from a few k to over 10k, and that was reasonable given the pain level.

However, keeping in mind this thread is about engagement levels, site layouts, etc, in my opinion, these low budget sites are irrelevant, their owners are irrelevant, and occupy the same niche that in the previous era 'brochure sites' did, those were sites you would whip up using any tool, and then popup live for a few hundred dollars following templates. Maybe a few thousand max.

layouts that drop user engagement though... it's hard to come up with any excuse for those.

Users being able to edit their own sites is critical in ALL cases, not just sub average, the days of my doing text and page updates are long, long, gone, and hopefully except for my own sites, will never reappear. Nothing is more boring than that.

ergophobe has the designer thing exactly right. When I'm trying to explain this to people I use the analogy of a contractor making the house. The better designers are aware of the basics of the construction issues and will not design something that conflicts with the building's requirements. But expecting designers to be good at information architecture is a reach. So it's best to determine the rough site structure first, then hand the designer a very fixed and concrete set of guidelines that determine what can be done and not done, particularly on technical levels. That works very well. A good design costs a few k I think, roughly, though it can go way way up, but it's really a total waste of money now if you are going the boxy route.

By the way, [velonews.competitor.com ] is a good example of information rich, yet very well done responsive layout. You'll notice on resizing that how well thought out the break points are, and what remains, and much information it contains. So there's no reason at all to accept the nonsense of endlessly flowing boxes filled with images being the only way to present adult non handicapped users information on mobile devices.

Re conversion rates, we made ONE single change in location to a core/key page component on a primary conversion page, and our conversions went up 100% almost immediately. That change was so obvious once we'd done it that we wondered why we'd never thought of it before. The 'experts' and 'consultants' had yielded an absurdly hard to navigate series of steps, that we have largely totally dumped in the process of refining the conversions.

While big sites like amazon might have to fight for 0.5% improvement, my feeling is smaller businesses can easily hit far greater than 100% increases if they stop listening to alleged 'experts' and start taking the question seriously, with real testing and real a/b analysis etc. We realized this when we came across an article on ab testing done by a guy for his web hosting site, just reading his methodology showed us exactly how to start testing various options. And it also made us realize just how important that initial first impression is, you can NEVER recover from it if it was bad because they leave.

But I also know that because each demographic is so radically different in mindset and how they perceive. Think of the following apparently similar groups: Big 5 sporting goods (a generic budget largely junk camping and outdoor gear store); REI (mostly yuppie wanna be campers and outdoors people); TarpTent ( a great cottage industry light backpacking tent maker); and Z-Packs (an elite, extremely expensive, exotic ultralight gear maker). Do you think that any consultant could actually understand what each of their markets wants, how they think, what they want to see? I don't.

A lot of these tools are really easy to add, heatmap stuff you can just pay for the service and you get wonderful heat maps by popping in a block of code, you only need to let it run a week, user behavior averages out, YOUR users, that is. Then you have to accept that what it shows you is true. a/b testing is harder, but it's not hard, the trick there is to show the a/b data alternating at the same time, because user behavior alters based on day of week or time of day.

None of this is hard to do, you just have to have the patience and discipline to do it, and to pay attention to the results. Note that comparing two crappy images, layouts, or text blocks, etc, with a b tests is not nearly as effective as testing significantly different items against each other.

For example, in the starting article, gray and weak and hard to read as it is, they did their tests wrong, they should have tested a GOOD vertical layout against the 3 column verticle layout, then tested the better of those two against the horizontal layout. On desktops it's almost a pointless test because the magic triangle in the above fold screen has been tested so many times it really needs no further work.
6:42 am on Oct 21, 2015 (gmt 0)

Preferred Member from BG 

Top Contributors Of The Month

joined:Aug 11, 2014
posts:547
votes: 174


I added heatmap to our website two years ago. Turns out our landing pages needed less text, as users who wanted to convert, didn`t need all the explanation. Rather they needed a peace of mind that our text content was properly written and with enough trust signals, for them to fill the (sadly huge but necessary) request form. As for layout A/B testing. I believe that before you even start meddling with different layouts, you should ask yourself - "Do I really care about mobile that much?" . Too many people obsess by the prospect of mobile, but we are long way from the desktop conversions and engagement levels and unless you are a service or product that can be consumed via smartphones, I`d relocate my design efforts in a different direction. As an example I really laughed out loud when I read that the online hotel booking industry was scoring BIG on mobile in 2014. In reality compared to desktop and e-mail conversions, mobile accounted for a nod of appreciation more than anything else. And do you know why? Because no person in his right mind would spend a few thousand Dollars or Euro for a vacation out of their phone.
3:14 pm on Oct 21, 2015 (gmt 0)

Senior Member

WebmasterWorld Senior Member ergophobe is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Apr 25, 2002
posts:8637
votes: 283


ergophobe has it fairly right


Yes, age and experience pays off again. In another ten years I'll get things mostly right. ;-)
7:47 pm on Oct 21, 2015 (gmt 0)

Junior Member

10+ Year Member

joined:Nov 9, 2004
posts:137
votes: 13


I had a further thought, and this is more in the area of ethics than anything related to web development, but it's my view that people who toss out wordpress junk with the 116 themes/extensions etc, that almost certainly MUST fail on a future upgrade, and, remembering, a non updated wordpress is a hacked server in the future, and wordpress requires VERY frequent updates, which should be applied within a day of the security update release, that combination in my opinion is web developer malpractice, pure and simple, and should be considered as fraud, unless the client is clearly warned about the long term risks of such strategies. Wordpress itself here is not the issue, it's actually the easiest to upgrade software package I know of for web sites, IF and ONLY IF the clutter is kept to a minimum, toxic themes are avoided, and clean templates are used.

This came to my mind because of an old client I made a custom site for, probably my last non cms type site I'll ever do that for, which is updated via flat data files, but the reason they picked that option was that I explained the requirement for a lifetime of future security updates vs a one time, clean and simple ball of code that would never require updating, barring trivial php deprecation fixes.

Nutterum, I completely agree, it's your demographic and user behavior that should dictate such decisions, not some random google babble spewed out by their script kiddie kids, who masquerade as web experts (ie, their current statement: design for mobile first). Our conversions are at least 90% non mobile, though what made me finally implement responsive, sort of, layouts, was not because of mobile phones, but because of tablets, which are in fact forming an increasingly significant part of our traffic, and they do convert. So do phones, but not nearly as much as non phones. But tablets do not require anything special as long as their viewport fits out site's width, which most do.

Re a/b testing, full layout testing is really hard, and should be carefully considered before it's used, but for smaller changes, it can be priceless. We've over and over seen our own assumptions be proved wrong by our users through a/b tests. a/b test data however has to be very carefully handled, a week is the minimum test time I find for valid data, and the overall numbers have to be fairly high (many thousands per a and b) before they settle down from statistical noise and deviations from the mean, which are also tricky to grasp empirically.

However, my guess is if a true ab test between a well done information dense layout and the endless flowing horizontal layout were done, the box layout would almost never be selected, for mobile or desktop. But my guess is also that this is never done. And if it's done, the results are ignored because of ego, laziness, or other non technical factors.
7:39 am on Oct 22, 2015 (gmt 0)

Preferred Member from BG 

Top Contributors Of The Month

joined:Aug 11, 2014
posts:547
votes: 174


The perception of these types of A/B tests in my opinion is flawed from the get go. The article presented by the OP was boiled down from the debate back in early 2014 when some big websites switched to "app like" UIs on their websites. Truth is Steve Jobs was right. Apps are better than search. And if done right people won't care about mobile browsing and use them instead. That combined with the huge amount of mobile users forced several key global businesses to "think mobile". They botched the idea because they were driven by hysteria that their companies will loose revenue from mobile.

Long story short - don't fall in the pitfall of layout game. Instead focus on many small a/b test UI and UX iterations until you reach a level of success above the industry average. Undertake heavy design changes only if your brand looks old.

P.S. Hell, I remember when I was shocked on a daily basis of the poor visual quality of the majority of the websites in the US. Then I realised that these websites were build to be functional and to generate leads not be an candy wrapper with no goodies inside.
7:46 am on Oct 22, 2015 (gmt 0)

Senior Member from US 

WebmasterWorld Senior Member tangor is a WebmasterWorld Top Contributor of All Time 10+ Year Member Top Contributors Of The Month

joined:Nov 29, 2005
posts:9927
votes: 974


Then I realised that these websites were build to be functional and to generate leads not be an candy wrapper with no goodies inside.


Hoorah! Thumbs up! This remains true to this day, and if one just makes the effort, still works with RWD... just a bit more thought into how the page breaks by viewport/zoom. Always the content. Always the user. Can't go wrong with that!
8:20 pm on Oct 27, 2015 (gmt 0)

Administrator

WebmasterWorld Administrator rogerd is a WebmasterWorld Top Contributor of All Time 10+ Year Member

joined:Aug 2, 2000
posts: 9687
votes: 1


Engine, thanks for sharing the post from my Neuromarketing site.

To me, the big takeaway is that some "horizontal" sites offer no clue that you have to scroll for more content. While most of the illustrations were from visually complex sites, to me the worst offenders are the minimalist sites that begin with a full-screen picture and some kind of short message. Visually attractive, but if there's content below that, like navigation, articles, or product links, etc. many visitors won't see it.

Good point about checking multiple devices, lizardx. I've actually darkened the font quite a bit from the original theme design, and to me it looks about the same on my MacbookPro, Dell laptop, iPad, and Note 3 (Android). I should probably darken it some more. Low contrast type reduces cognitive fluency - unconsciously, sometimes, but our brains find it harder to read. That, in turn, may cause a visitor to stop reading sooner or be less likely to take a requested action. (The worst offender I can think of is WIRED magazine - they frequently print tiny fonts on colored backgrounds with low absolute contrast levels. Ugh.)
7:56 am on Oct 28, 2015 (gmt 0)

Preferred Member from BG 

Top Contributors Of The Month

joined:Aug 11, 2014
posts:547
votes: 174


@rogerd - your post about contrast and readability put a smile to my face. When talking about the topic with peers or clients I always refer them to SAP. They employ contrast ratios of 4.6 to 1 and even though some of their products/website landing pages might look like they were dug up from 1998 they are easy to read, easy to understand and easy to promote action taking. I always hate when some newly self-proclaimed designer tells me that certain layout looks perfect on his 29 inch thunderbolt displays, neglecting the fact that the vast majority of users use screens that are nowhere near that.

@tangor - Yes. I was was young and wanted to make the internet a beautiful place. A few years passed and I looked at my office window. I saw many ugly looking buildings that were functionally perfect :) . We live, we learn.
This 59 message thread spans 2 pages: 59