While that may have been true in the short term, there is a perception now is that this may have been when eBay jumped the shark. The “buy it now” button, while having that immediate effect on auctions, also had the unintended (?) consequence of allowing people to force a storefront square-peg through an auction-site round hole. Many auctions look like an online shopping mall, where the auction-dynamics are overwhelmed by the mall-perception.

Will eBay win versus craigslist for people who still want that “auction experience?”

Do you see this design-ROI discussion purely as “payback for art” aka “make it pretty and I’ll give you a dollar?” I ask because it seems like the argument you’re making is that “making it pretty has value, but that value is subordinate to everything else.”

If so, then totally my bad – I’m really trying to talk about all of the elements of design from aesthetics to affordances to information architecture to usability. I was trying to make a much bigger point, and I guess I didn’t.

If you could depreciate products on the basis of the last time they differentiated themselves and the length of time that the differentiation held or earned a premium, you could easily see that other offer elements like design were earning their keep.

Return on design is obvious in the economically indifferent outcomes. If you are earning more since you went art, then blame it on art. It’s pretty hard for some founders, if they are still around, to let art be the winner, but it is a phase specific issue. It is not always the answer. Nor, does it necessarily hold for long.

The brand dependent company is also promo spend dependent. Eventually, even art is copied, so it ceases to differentiate.

Thanks for reading and commenting. Great point about A/B testing. That was one thing I always felt like Amazon did correctly. When rolling out changes to their site, they would randomly route select customers to the new version of the software – allowing them to track specific statistics, like how many people complete transactions (measuring the funnel). Or if version X of the cross-sell design is more profitable than version Y.

When contracting for copywriting for an ad, a company may very well hire multiple copywriters. They then would place the ads wherever and see which one got more of the desired result.

That ad becomes the “control” and they constantly test new copywriters’ ads against the “control”, always seeking to get better response.

With modern technology, companies can do a simple A/B Split on web traffic and test out two (or more) different designs to see which one has the most of the desired effect (more usage, more traffic, more purchases, more requests, whatever).

Dan