Questions

What are the best ways to test 2 different version of product page/user flow?

Hello all, we are building a website which is something like groupon for fashion. We have 2 product pages in mind. What's the best way to test, learn and improve? Anyone has experience on this?

4answers

Easiest way would be to use an existing A/B testing service (e.g. https://www.optimizely.com/)

Or, you could do it manually by sending out (via email or whatever) different links to different people.


Answered 8 years ago

A couple of great services that can help you out with this:

www.optimizely.com
www.wvo.com


Answered 8 years ago

I have experience at several companies testing website variations. What you are looking to do is commonly called split or A/B testing. There are several companies that provide split testing services.

My favorite testing tool is called Optimizely. (https://www.optimizely.com/plans/#testing). It has free and paid plans depending on your needs. The tool has a lot of engineering going into it, meaning it probably won't mess with any other analytic services you might have on your site.

I have used split testing solutions that caused Google Analytics and other tools to incorrectly pick up data. If you are using any scripts or third party services on your site, it is very important to make sure split tests do not interfere with them.

Optimizely will allow you to test different pages or even minor alterations to the same page. You can make these modifications using a powerful editor that requires little coding knowledge.

Before you setup a test, you must make sure Optimizely is tracking the right goals. Goals are any interactions that indicate the test was a success or failure. You can track specific clicks on the page and whether other pages were visited.

After enough people have visited your test, you may reach statistical significance. Statistical significance means the sample size of people in the test was large enough to be reasonably sure the winner didn't win by pure chance.

The hard part about split testing comes when you can't reach statistical significance or can't make sense of your results. That is where a conversion rate optimization expert or growth engineer can help give advice on next steps.


Answered 8 years ago

You have some options and your choice depends on what point in the design process you want to test it.
Option 1: create a prototype of each and do a moderated or unmoderated user test study with it.
Option 2: Develop and launch both versions and do it as an A/B test where half your users are shown version A and half are shown B. Then you look at the data to see which version converts better.


Answered 5 months ago

Unlock Startups Unlimited

Access 20,000+ Startup Experts, 650+ masterclass videos, 1,000+ in-depth guides, and all the software tools you need to launch and grow quickly.

Already a member? Sign in

Copyright © 2024 Startups.com LLC. All rights reserved.