If you need to do some A/B testing at minimal cost, this is our quick overview of the available options that we found.
There are some powerful free options, and some free trials for some superb products. As with all things, the free options come with limitations that mean that for some circumstances it may be totally worth paying up for the right product/service. Free is not always better, and paid is not always better! We leave the final decision and hard negotiation to you :)
If you're running a big site with budget, this is great. You might be able to get the results you need from a 30 day trial, but they clearly hope you'll like the experience so much that you'll stay.
There is a Drupal module, which helps you add the Optimizely code to your site and control some of its actions. It's not in active development, but has Security Coverage.
Visual Website Optimizer has a 30 day trial, then paid plans. No public pricing.
VWO is much like Optimizely, and has a bunch of really nice features that make it worth doing if you have some cash available. 30 days will only get you so much data, but if you just need a couple of quick tests, it might be worth trying out.
ABJS is a free contrib Drupal module.
You'll need to wire in your analytics events tracking yourself if you need it, but that's not too scary for most developers. Google Analytics is generally quite a dream to work with nowadays. Plus, we'll have a guide to setting that up later in this series.
Google: Content Experiments
ContentExperiments is a free offering from Google Analytics.
They're having a really strong go at making the GA platform provide everything you could ever need, and this is a hugely powerful set of tools.
With Content Experiments, you need to create the page variations yourself, then tell Experiments about those variations so that it can work its magic. This is a slightly different way of working, so might be a bit of fuss. But there's probably a way to make it work with Tag Manager and some JS if you want to. As always, Google have made an interface that's nice to work with.
The most interesting tool is the Multi-armed bandit experiment, which can take a number of page variations and rapidly decide which is the most effective for the goals you've specified. This looks to be really useful if you want to compare a selection of different layouts.