ThePathToFreedom

It's been awhile since I've vetted the options.

Anything new I should know about for split testing?

What are you currently using for split testing? And why?

Thanks

richpage75

Great question There are two great low cost tools for doing A/B testing:

- Visual Website Optimizer. These guys have been around for a while now, and are a great visual editor that allows anyone to create tests with no coding skills required. They also offer some great functionality to help you improve CRO - built in surveying tools and clickmap tools to help you figure out what to test. I really like their 'idea factory' too.

- Optimizely. The new kids on the block - actually cheaper than Visual Website Optimizer, but charge more for their advanced features. I actually prefer this tool, as it offers superior functionality for advanced testers like myself (like cross-page testing).

I would forget about Google Content Experiments (which used to be called Google Website Optimizer), as it has really poor functionality, and just allows you to test whole pages with a basic redirect, not specific page elements only. A real step backward since they changed the name of the tool.

If you would like to see more, I created an article that compares these A/B testing tools on my blog. If you would like to see the URL, let me know. Thanks!

Rich Page
http://www.rich-page.com | Twitter: @richpage
Author of 'Website Optimization: An Hour a Day'

stopper

Thank you richpage for that great info I will be sure to check it out. How can I get the link?

http://www.charlesmomo.com Are you interested in 200 to 400 visitors a day to your site

richpage75

No problem, glad to help. Here is the link for the reviews of the tools:
Visual Website Optimizer: How it Compares to The Rivals

Rich Page
http://www.rich-page.com | Twitter: @richpage
Author of 'Website Optimization: An Hour a Day'

high_plains_drifter

both too expensive for my budget, I wish there were smaller packages available for people just starting out, im a firm believer in constantly testing and tweaking

Want to know why your website is NOT converting? Have a UX expert review your site for just $29

webtraf101

I also recommend Optimizely as an split testing provider which was founded by two former Google product managers, also it allows you to target split testing variations to visitors based on traffic source, language, cookie, or browser allowing greater segmentation of your audience. It will lets you create unlimited variations and is extremely simple to implement with both HTML/Javascript editing inside the platform itself or the ability to redirect to an already created variation. The platform allows unlimited goals with real time results available immediately. Some of the other features Optimizely offers include mobile website testing, cross-browser testing, multi-page funnel testing, multivariate testing, geotargeting & IP address targeting.

Mrsparrow

Sorry for the blatant self-promo here but you can also try Marketizator. Besides AB and MVT testing, you also have the possibility to run online surveys and triggered interactions (on load, on exit & on scroll popups). We have really cool segmentation options such as location, weather, keyword, traffic source or history + many other features.

dburk

Hi high_plains_drifter,

If you can make your own web pages then you can use the Content Experiment feature of Google Analytics, it's free and can delivery worthwhile benefits from A/B split testing. While it is not as convenient as other packages, and it requires you to have the knowledge to create your own web page variants, it can be a good option for someone on a very limited budget.


Hi richpage75,

In case you weren't aware, you actually can use GA to do testing on specific elements in much the same way that other tools do that. The downside is that it take a bit more time and knowledge to setup Content Experiments in this way. Other programs make it relatively easy with editing tools. If you can make your own website code changes then you can use GA Content Experiments to test those individual elements without a redirect script that points to a different page.

See how: https://developers.google.com/analyt...ts-client-side

And for programmers that want to setup Server-side experiments using Content Experiments you can do that as well.

See how:https://developers.google.com/analyt...ts-server-side

* Get Results - Outsource Your PPC Management |
* Don Burk Advertising & Marketing - www.donburk.com

Amod Oke

$49/month expensive for you when you want to constantly be better with testing and tweaking?

If you cant afford 50 bucks/month to split test, you shouldn't be split testing.

crams

You can start building multiple LP's and use bevo LP rotator to split test your lander.

savidge4

Aside from the tools themselves... indirectly a point has been made through this discussion about the expense of testing. I am going to take that a step further and assume that they are obviously not making enough money which in turn means they are not getting enough traffic.

So based on an assumption that traffic is at a minimum.. what level of traffic is needed to get a "Decent" A / B test result? For me personally I have sites that range from 1000 visitors a month up to 30,000 a month, and one site that is well into 6 figures.

I have been playing around with "Short Testing" run 1000 through "A" and then 1000 through "B", and then "C" and "D" etc to close in the options that I want to "Long Test". The Numbers for the "Long Test" are more or less dictated by the amount of traffic the site gets. I would say and average for me personally is 5000 or so.

I know there is not a magic number, but when is enough enough?

Mrsparrow

Here's a more elaborated answer to your question: http://www.qubitproducts.com/sites/d...e_illusory.pdf

Also when running an AB test you have to take into consideration some other aspects such as:
- think of business cycles (weeks, months, seasons, etc) You should test at least 2 complete cycles if you want to get relevant results
- try to avoid simultaneous AB tests
- aim for statistical confidence rates higher than 95%
- don't stop the test just because the results confirm your hypotheses

Marty Foley

Here's a comparison of a couple dozen or so A/B and multivariate testing tools:
Which split-testing software should you use? | Conversion Rate Experts

Personally, my two favorites are Convert.Com and VWO.Com

KyleChapman

We really like VWO and Test and Target. Those 2 will take care of most of what you need.

However I've used Optimizely, Google's testing platform, and many others.

It really depends on what you're looking for. It's like buying a car

I've done conversion work for many big companies such as Expedia, The Boston Globe, Kobo Books, Firestone and more...

Kyle Chapman - CEO - Conversion Marketing Services a Conversion Rate Optimization Firm