Comments on: 4 Reasons for Site Testing: Optimization Testing 101

By: Linda Bustos

Linda Bustos — Thu, 29 Jul 2010 00:11:18 +0000

Hi Alvin, Part 4 of our series is about coming up with a testing process that includes a bit on deciding what elements to test
http://www.getelastic.com/developing-a-site-testing-process/

My philosophy is you should begin with web analytics to understand as much as you can about customer behavior (segment, segment and segment some more) to help you shape your approach of what to test. User testing can also expose some of the bigger problems on your site. The rest is more art than science – you need to apply common sense to testing, but even the best common sense can be wrong – as many have found out our guts can fail us sometimes! There are no right or wrong answers about what to test. Every situation is unique.

I do agree with you that there are minutiae that sometimes people get sidetracked on. For example, testing cart button colors and designs when the bigger issue is poor product description content. Let’s get some more ideas going in the comment thread…

By: Audio Bible

Audio Bible — Wed, 28 Jul 2010 22:23:40 +0000

I have gotten into testing in the last year. Biggest thing I can suggest is get someone good to set up yout Google Analytics code on all page and checkout process, if you use Paypal as well it is a bit tricky to do. I used Blast Media, they are very good in my opinion. If your analytics are not right on, your numbers are not going to be 100% correct.

I learned this after paying someone to help with optomization services, one that is highly well known. they did not have a GA setup expert, so we were never tracking with good numbers.

It is worth the little bit of money to get the GA setup correctly first. you will spend much more than that testing and optomization services….

By: Christina

Christina — Mon, 26 Jul 2010 04:47:13 +0000

I’ve never done site testing before but now that I’ve read your blog, it is great idea indeed. Thanks.

By: Alvin Tan

Alvin Tan — Sat, 24 Jul 2010 12:20:05 +0000

I think that testing, tracking, and measuring conversions have become mainstream ways of optimizing a website, and there are a ton of resources out there addressing the “how” of site optimization. What is less covered is the “what” of site optimization — specifically what hypotheses to pit against each other, and how to generate those hypotheses. Generating the wrong ideas or hypotheses causes time wastage and causes marketers to test a whole bunch of sub-par site versions, none of which are even near to “optimal” state. There could be a million reasons that a website is not converting well and I think that the most valuable complementary skill to knowing how to test is knowing what to test.

By: Kenneth Dreyer

Kenneth Dreyer — Fri, 23 Jul 2010 16:07:07 +0000

Good article! I look forward to see what more you come up with in this series.