The Only Button Design Software You'll Ever Need...
Split Testing For Profits.
What exactly is 'split testing', and why is it so important to my
Split testing is a process - these days generally an automated one - of establishing the
individual components within your website design responsible for generating conversions and
It's about taking individual elements - headlines, copy, calls-to-action and graphics generally
- and testing each in isolation, in the hope of ensuring (ultimately) each and every
component on the page is pulling its weight as effectively as possible.
Split Testing Example:
By way of an example, you could test six or seven
completely different headlines against one another to establish, definitively, which specific
headline leads to the most conversions (those conversions being visitors taking the desired
action on the page in question - buying your product for example).
Having established which is the best, you would then make that headline a permanent
fixture, and move the focus of your testing elsewhere.
Systematically, over time, your testing will reveal (in the real world environment) what works
best on your site, in your niche, with your visitors.
There's just no substitute for this kind of empirical data. Without the evidence - the
proof positive - you're leaving an awful lot to to chance and guesswork.
You’re going to need four things to get going with your split testing:
Something to test (e.g. headline, main graphic, call-to-action, etc.)
Some variations of the test element (so, different headlines, alternative button
A tool for the testing process (more on that in a moment)
Traffic (visitors to your site)
The first two items in the list are easy enough to sort out - you just need to start at what you
feel are the most important page elements, and work from there. Coming up with variations is
simply a case of brainstorming some creative ideas.
For brevity's sake I'm not going to write a step-by-step guide to using this powerful tool here
in this e-mail - I'll try and put something together in the coming weeks to help you get started
using it - but in the meantime, here's a link to to a pretty succinct explanation of how it works,
from Google themselves:
The final prerequisite, traffic, is of course something you’re going to need to work on (if you
don’t already have plenty of it, of course!). The reason you’ll need a reasonable level of traffic
is that Google’s testing tool shares that traffic evenly between the test variants, then monitors
the conversions resulting from each.
For the results to be statistically valid you’ll need to send hundreds (but preferably
thousands) of visitors through each of the test variants. So, if you’re testing several variants
you’re going to need rather a lot of traffic (else the test will take a long time to
In case you’re in need of more traffic, there are some tips/techniques over at ShoutMembers and
we’re also working on some 'mini-courses' at ShoutTraining
for you, in case you’re interested.
Tip #1:Keep your initial tests small-scale
(ideally one variation compared with the original). This way the test will complete sooner rather
Tip #2:Start with the main page elements, such
as headlines, calls-to-action, price and so on. These are some of the ‘pinch points’ on your site
where visitors will make snap decisions regarding whether to buy or not. These are where the
biggest gains (or losses) will be made in your testing, so it makes sense to start there.
Tip #3:Make your test variation significantly
different to the original. If your test variant is too close to the original it will take
Google longer to identify a statistically valid result.
Tip #4:Record the results. I find it’s a good
idea to keep a spreadsheet of the test results for future reference. The results you’ll get will
tend to be fairly ‘statistical’ so it makes sense to write a short, plain-English
explanation of what the test was about, and what the results mean, so when you come back to
the list of tests at a later date you will be able to quickly decipher the results from your
different tests, and trends will be easier to spot, too.
Tip #5:Get on with it. Split
testing might seem like a complicated, technical process but believe me: once you’ve done a few
simple tests you’ll find they’re easy and quick to set up, so you really need to dive in there and
have a go.
As I said at the top of this message you really can’t afford not to run these essential
I hope this short guide to split testing helps you understand the process, and that the video
helps you get to grips with Website Optimizer. Wishing you every success with your tests!
Really simple to use - I designed a whole set of buttons for my
site in less than 10 minutes.
Sally Van Es
Definitely worth ordering at such a low price!
Mike Ramshaw, UK
The perfect addition to anyone's designing toolbox.
About ShoutDesigner Button Creator Software.
Button Creator tools don't come better than ShoutDesigner.
With a great-looking interface, world-beating quality button templates (all fully-customizable),
ShoutDesigner will have you creating professional-quality button graphics in minutes, regardless of
your design ability.
The free version includes 50 button styles, and a 'pro' version contains 200 styles and extra
features (only $27). Truly the only button design software you'll ever need.