Dealing with some designers is hell. I’m talking about designers who think they know what a good design is and blather on about their superior knowledge. They don’t know what good design is. There’s no way they could. They might have a better idea of what might work, but it’s just a guess.

I’m lucky enough to work with a great designer — Rachel Cunliffe of Crea8d Design is a New Zealand National Treasure.

Most designers think they can just look at a design and tell whether it’s good or not. In fact, all they can really tell is whether they think it looks good. Until a site is actually used, and can be shown to function correctly for the purpose it was built, no-one can anything except that they like it or not — even the designer.

That’s what’s so galling about the tawdry “critique” going on over at Business Logs (and elsewhere). Rachel, who did this Names@Work site, our Pulse site, Darren Rouse’s ProBlogger, and many others, recently redesigned Mike Arrington’s TechCrunch, and she’s catching a load of really undeserved garbage.

Mike Rundle (a competitor no less) has decided to pontificate about what works in a design and what doesn’t. He goes into things such as the proper angle from which light should appear to emanate when doing gradients; what Web 2.0 is and isn’t (according to him, it’s rounded corners and a particular typographic style); how the previous design was somehow easier to copy for clients who came asking for that cool Web 2.0 style.

He quotes other people who have similarly well-informed insightful comments — for instance: “NEVER display ‘Your ad here’, very tacky.” He quotes them as if they’re evidence of some kind.

Speaking from the client point of view, if I have evidence that “Your ad here” is attractive to the people who use my site, you can be sure that I’ll put it up, because I don’t have to go to cocktail parties where snarky designers decide what’s tacky or not.

The only criterion worth measuring a design by is whether it works. Designers can drive customers away from their own sites, but if it’s my site being designed, I’m going to insist on something more substantial than rounded corners, pastels, and and spurious advice about how to do a Web 2.0 look-alike site.

When it comes to a critique, any criticism of a site that aspires to be anything more than uninformed opinion has to cite some kind of evidence that comes from people who use the site — abandonment rates, ratios of visits to page views, subscriptions per visitor, and so on. Ideally, this evidence is tied directly to the goals that the business has for itself, whether this is increased sales, subscriptions, submission of photos, whatever it is. An initial design is a starting point, and it needs to respond to the people who use the site and change according to their feedback.

I want my site to accomplish its mission, and so there’s nothing that annoys me more than this sort of arrant nonsense from Mike Rundle:

… although Rachel had her thoughts and ideas about what a new TechCrunch should look like, they were pre-empted by what Mike gave her as an instruction set. This happens a lot in client work, which I’m sure Rachel knows very well, and that’s the idea that no matter what vision you have in your head for a client’s site, it doesn’t matter if it doesn’t match their vision for the site. In this case, I’m sure Rachel had a great design all dreamed up and then Mike’s requirements were the total opposite. That’s really tough.

[Translation: Clients are idiots -- if only these morons who pay us would get out of the way and let our vision shine through. I imperially forgive Rachel for not being strong enough to resist.]

Since we’re all just giving opinions here, let me say that Mike Rundle’s “critique”, and the attitude that accompanies it, is a huge pile of horseshit.

Listen up, designers — if you’re really interested in Web 2.0, it’s not about light from the top left, it’s not about rounded corners, it’s about the people who use the site. It’s about communicating with them. It’s about putting the power in their hands. That means that you make your best guess about what works for them, then you get feedback (often by measurement), then you refine. Rinse, repeat. And you keep on doing this, forever and forever. You do NOT, ever, put out a perfect design — because it never is. You work with your client, not in spite of your client.

Working with Rachel is radically different than working with the snark crew. Instead of treating me like an idiot who needs to be put up with, she acts as a partner. Instead of spouting airy nonsense about rounded corners, she’s interested in knowing how I think the site will be used. Instead of quoting dogma about where light should appear to appear from, we talk about what people are likely to want out of the site and if we’re giving it to them. When the initial design comes out, we listen and learn, and Rachel makes improvements based on feedback. Now that’s a designer. Mike Arrington is very lucky to be working with her; I hope he knows it.

This kerfuffle must be causing pain to a very professional, very classy woman who takes her work seriously. It’s mean, it’s cheap, and worst of all, it’s bad and baseless criticism, a collection of ill-informed opinion, with no reference at all to what makes a good design.

Which just goes to prove the old saying that there are more horse’s asses mouths than there are horses.

UPDATE. Rachel has resigned from the Techcrunch blog after a what I can only assume is a lapse of judgment by Mike Arrington in publishing someone else’s idea of an improvement to the design. Good for me — now maybe she’ll have more time for the projects I want to hire her for. Also, Darren Rouse adds some of his usual temperate good sense over at Problogger.

Tags: design, web design, web 2.0, Rachel Cunliffe, Cre8d Design, Michael Arrington, Techcrunch, Mike Rundle, Businesslogs, completely classless

Posted by Antony in: General