Home, Blog index

That old design v content chestnut
2010-06-22: I am trained to act like an engineer: only to do what is proven to work. I use science and knowledge to cut through the Internet marketing bs and seek out the truth .. what works.
One of the things I hold true is that typically a client puts far more emphasis on graphic design than the facts support. So, for instance, if a website costs £1,000 the client's requirements for the site to look great will take up maybe £500 of that, leaving only £500 for things like content and functionality, not to mention usability, noteworthiness, functionality, usefulness, accessibility, internationalisation, ease of future maintenance, good testing, security, findability, conversion, persuasion, and information architecture.
I've done long development projects where I sit with the client and we talk about the sort of functionality the site should have and what data it should store, and after a few months of fair to middling disinterest, the day I turn up with a rough idea of what the site might look like, the client gets all animated: can we put this here and that there and can I have it in Everton blue because I support Everton?
Almost every time, the graphic design input from the client is simply based on personal preference .. what they like. A couple of decades ago the old designer dad of a friend used to say "I haven't designed it so you'd like it, I've designed it so your prospects will like it".
I might not be a graphic designer, but I do know something about what graphic design needs to achieve, about branding, about guiding the eye, about associations. Those things might be accepted intellectually, but the old brain just overrides all that and before you know it, you're back to what the client likes. I do think that's it .. there's an innate part of all of us that almost can't help but get involved in how a site looks, no matter how much intellectualising we do about it.
It also feels like it's also about the client wanting to be involved and to contribute, and with the best will in the world, talking about databases is hard work. It's hard to visualise the impact of any decision. By the time it comes to how it looks, all that pent-up frustration comes out: "At last, I can contribute, brilliant, I'll have that there, and a border around that .. "
So I've been working on ways around that, ways not to trigger that 'personal preference' thing. Wireframes and mood boards and all that. Ways to get consensus and signoff on the look of a site.
Anyway, I don't know if you can see this discussion in LinkedIn about the relative merits of content and graphics. A chap called Brandt Dainow says "I do analysis of e-commerce sites for a living. Unpopular as the opinion is, I have yet to see any data which shows site performance is affected by graphics, unless the graphics are so bad the site is unworkable. Statistically, the highest conversion rates are achieved by pages which contain 471 words of readable copy. Graphics are essential if you're selling products, in which case you want 3 images per product. Graphics are important if the topic is visual - like a picture or a video. But in terms of general design principles, the less artwork the better. That's not my "opinion" that's what the numbers say. People won't be convinced to buy something because you have a nice logo or color scheme - it's the words which convince."
I just want to hug him.
Anyway, the point is I just checked the results of a test I did on personalisation. I installed a word cloud that provided navigation to the most popular pages on a site, and then I made it so that half the effect came from the pages that were popular with that particular user.
It's a very basic form of what the BBC called in one of their website iterations "wearing a path", so it becomes easier to find the pages you've visited before.
Anyway, with 6 months of data in, I see a 30% increase in conversion. I've never seen that from a graphic redesign.

By John Allsopp
Next: A competition flash - Previous: QR Code - Back to blog list

Tweet this page Tweet this blog Share this blog Share Bookmark this blog Bookmark

blog comments powered by Disqus