eCommerce Design

[Lao]

Why take a punt on 1 design? Create several homepages, split traffic equally between each variation using Google Optimizer (for example) and see which one the users actually prefer. That way you're letting the most important people decide rather than the vanity of the designer, or of course marketing experts.

Good idea, but one problem may be the confusion it would create for returning customers. Consumers prefer familiarity, and if a site they were happy with first time has changed drastically they could be put off, even more so if they visit regularly and see several versions of the site.

This is a really good site http://www.archive.org/. Type a web address into the 'Wayback Machine' and see what a site was like several years ago. Good for seeing how they have matured and developed with technological enhancement and changing consumer habits.

Look at the original Tesco or ASDA sites, or the EFC site from 1998 for example!!

 

[atrottel]

Good idea, but one problem may be the confusion it would create for returning customers. Consumers prefer familiarity...

Not so much in real life. Wholesale site changes and testing tend to get done before they go live. Conversion rate optimization is generally dealing with small changes which can make a big difference (such as text content or position of buttons in shopping cart), but that which you wouldn't really notice (apart maybe from saying that a site has got easier to use over the last year).

Compared to the general content changes that (should) take place on most sites, the split testing shouldn't be a big deal.

There is also an argument that says that familiarity leads to site blindness over time.

or of course marketing experts.

Of course, if they are marketing experts, they will be using CRO. The beauty of tools like that is that they are measurable (do I have more visitors?; do they convert?; are the changes profitable in whatever parameters I am measuring?).

Take GOT as an example of a non-e-commerce site:

Another Everton site goes down. D/D get a load more visitors. Do they participate or donate (conversion)? Do they change the site for good or bad? Does the advantage of their participation outweigh the extra bandwidth or other server costs (profit)?