The Three Sins of Graphic Design
Not thinking CMS
Unless it's a trivial project, there will probably be some kind of content management system involved. This means you don't KNOW how long the titles will be, how many paragraphs of text there will be or how long the page titles will be. Make sure you have the real site map and titles, that you explain the approximate maximum lengths for things like page and article titles and what will happen if something is too long to the content authors before you get sign off on a design.
Not thinking CSS
I lose count of the number of times I've seen cool graphic designs that are a pain to implement pixel perfect in CSS. Please learn what's easy to implement, what isn't and if you're tempted to show a client a design that is going to be hard to implement, get the client to sign a rider saying the following:
"I understand that by the time I'm done with a site template, 20 screen templates, any changes I request to the agreed design after implementation and then fixing those changes across all browsers, I'm probably going to end up blowing an extra couple of grand on this design rather than something very similar but easier to implement. I truly believe that having this precise layout is going to sell substantially more widgets or consulting hours than the almost identical but easier to implement alternative."
I'm not saying there are no use cases where a harder to implement design will genuinely blow conversions out of the water compared to a similar but simpler design and spending the same money on split testing the copy, offer and title, but I'd like to see the results of the testing. I think most of the time, a simpler design and more money to test and tweak the pitch is going to work better. But maybe that's just me - I ran an ad agency for 10 years, but I was a copy-writer, so what do I know about design?!
Not thinking IA
PLEASE, pretty please, don't "balance up" the home page with testimonials just because it looks a little dull and you have pictures of people handy. Get with the client, tell then you want to balance up the home page and based on their essential use cases, come up with the best use of the space that looks cool AND achieves their business outcome - even if it means you have to spring for some new images for the comps. Please . . .
Any other sins of graphic design(ers) I missed?
Good luck with that one.
Remember that these people are designers and not programmers. They don't care about CMS, CSS, IA or any other acronym you want to throw at them, they care how pretty it looks.
Thinking that you're going to train a designer to think about anything else is like training a zombie not to eat human flesh, it just ain't happening.
With that said, rather then getting all bent out of shape, learn to work with your designer. I always have my designers give me the PSD (Photoshop) or AI (Illustrator) files instead of giving me a Dreamwaver half ass attempt at HTML or a JPEG of what the page looks like. With the orginal PSD or AI files, I can recreate the layout in XHTML and CSS in a couple of hours and I don't have to sweat it or get stressed about any of it.
You willing to fixed bid that at two hours for quality, cross browser tested, best practice, well named, semantic XHTML and CSS?! If so, we should talk offline about getting you to do all of our psd to CSS work :->
Please clarify that includes any tweaks to the CSS required when the we notice that if titles are too long to fit onto one line they don't look good, if there are two many words in a teaser it breaks a layout in IE, etc.
If I could get that commitment for the cost of two hours from someone who is good at CSS and programming, my problems would be solved :->
Any medium-sized project requires a number of skills: graphic design, GUI/usability design, html/css, javascript, flash, actionscript, a server-side language (PHP/ASP/NET), database admin, SEO, maybe even 3D and some video editing. Obviously this requires a number of people to work on the project. If they don't work together, at least acknowledging and understanding the limitations, the project will not ultimately succeed.
Two examples: I am not a SEO expert, but I work with one and we take care of some layout considerations right in the beginning (for example column order).
Second, we have a brilliant flash animator; he does mostly character animation. It would be foolish to try to make him an actionscript coder, because he'd never be as efficient as a dedicated developer, but at the same time he has to understand streaming, file size optimizations and some basic commands.
As far as "Thinking CMS", I have to laugh. You are basically asking the designer to think in your world which is just as difficult as asking you to think in the Design world. If every site was designed to be easy to implement using a CMS, then every site would look the same.
Imagine that every magazine ad was based on the same template, or every billboard, or every TV commercial. Designers have been designing ads way longer that developers have been building web-sites. And thus have built up and expertise in creating the emotional response required to meet the business goals. As web developers we need to make sure we support that instead of forcing something "easy to implement" down everyone's throat.
I'm not saying that ALL designers don't care about IA. I'm saying that SOME designers don't care about IA. If you have never come across a designer more focused on look than usability then that is great, but unfortunately they do exist.
As for thinking CMS, I think it is a requirement that any competent designer be ale to create a design while thinking about the range of possible content rather than a specific piece of sample content. They should always think about maximum text lengths, how long titles might spill over, what will happen with longer content and negotiate these ranges with content authors. What is the point of a beautiful comp that looks lousy on the website because it doesn't work with real content?
I'm open to taking on sites that are not easy to implement. All I'm saying is that let's say with 20 screen templates, a site templates, various post implementation revisions and the usual problems with a design not working with real content, with custom CSS and HTML you're going to blow a non-trivial amount of additional time and money on the harder to implement system. If you can sell the client on an additional payment for that rather than for testing offers and the like (as direct mail professionals have been doing for a long time), I have no problem with that. I just want the client to sign a waiver saying "I get that it's gonna cost a couple of grand more by the time we're all done with this difficult to implement design and I believe that'll give me a business ROI greater than any competing use of the cash". However I see many small projects where an inordinate amount of time is spent on implementing a "difficult" design just because the designer didn't know easy from difficult. That isn't their fault, but I think we should tell designers upfront and give clients the option of a more skin driven approach or a completely custom approach so they can weigh the business cases.
2. Showing the client mockups before the developer sees them. This would always result in some weird design feature that the client loved but could not be replicated easily with CSS.
It all boils down to communication between the designer and developer. If each can learn a little about the design and development processes... things will go MUCH smoother.
don't code their work in well-formed XHTML (I love me some tables)
don't build their work to utilize pre-existing CSS styles (hmm... i'll just use a font tag to style this)
develop short-sightedly with no thought of future expansion (but the project scope didn't ask for that)
don't care to take the time to learn how working with a front-end developer can make their life easier (dude, I'm not a designer)
http://csssage.com/
http://www.psd2html.com/
http://www.psdxhtml.com/
http://www.xhtmlit.com/
:)
How do mean "pessimistic"? I've worked in this industry for over 14 years now and seen definitely more than a hand full of agencies on the inside. What I've noticed is that the web development agencies that are successful are the ones that are ran as software development companies where visual design is the support role.
It isn't CREATING the CSS that costs - it's the tweaking of it as the design meets the real world content.
Has nobody else ever had to spend any time tweaking their CSS when a stylesheet they were given looks great with original content but then becomes a problem for templates where a product has no images or too many, where page titles are too long, where article titles are too long, where a product doesn't have a price and it messes up the layout, etc?
Thoughts?
I personally use services of xhtml candy they deliver on time with w3c valid clean code.