Reply to thread

Derek,

I've done my homework on split testing, I follow the feeds of the best in class like Conversion Rate Optimization | Conversion Rate Experts and Tim Ash's SiteTuners | Website Conversion Rate Optimization ? Landing Page Testing » SiteTuners and others.  Optimizing any site requires many steps.

 

Off the top of my pointy head..

.

1. An indepth plunge into the shoppers world and then the same deep look into the seller's operations to discover needs and wants of both sides. Shopper surveys are needed. YOU CAN'T SKIP THIS STEP.
2. Build Creative to "bring to light" the discoveries found from the needs and wants of both sides
3. Examine the current website to find strengths and weaknesses (with heat maps and focus group users, etc...)
4. Formulate split test game plan
5. Execute split testing
6. Formulate multivariate game plan
7. Execute multivariate testing

Next, if you don't have enough traffic, your results could easily be flawed.

Split testing done right is lots of grunt work and is very expensive.  Next, the larger the sample, the stronger the rules become.

Bottom Line:  Split testing on a dealer level has a high potential of producing bad data. IMO,  Optimization to be done at the Vendor level.

See Uncle Joe's Rule #121:  "Just Because It's a Good Idea, Doesn't Mean It'll Work".