It may sound spooky, but running online marketing experiments on people isn’t really that bad.
[feature_block style=”icon” overall_style=”icon” columns=”1″ icon_style=”2″ font_color=”#000″][feature title=”” icon=”422.png” upload_icon=”” bg_color=”” href=””]The benefit of running an experiment is that a small test will show you what will do best, and then spending more money on the winner will get you a better profit or result. And with online marketing it’s quite fast – no waiting around for weeks until the results come in.
[/feature] [/feature_block]
An experiment helps you find which ad, page or set-of-page variations does best.
The most simple is split-testing where 2 variations of an ad, email or a web page design are run. There’s usually a single difference between them and they’re normally run simultaneously.
Then it gets more complicated with multivariate tests, where different headings, copy and forms in all their combinations are tested at once. And there’s also full funnel testing where you measure the impact of 2 competing series of pages.
But I find that if something’s too complicated, then it doesn’t get done.
And also, a simple A/B or spit test will often help you out more than the complicated tests just per the 80/20 rule.
Technical note: sometimes split-testing a page is called A/B testing, which refers to sending traffic to a single URL ( like www.domain-name/page-test.html ) and the 2 page variations are alternated using code. Split-tested pages each exist on their own URL.
How to run a split test experiment
Split testing ads is easiest and especially easy on Facebook. You simply publish 2 ads in the same ad set with a different picture or text. Set a low budget for a couple of days. Then pause the losing ad and increase the budget for the winner !
You can do the same in Google Adwords, their display network or any advertising platform.
then if you’re sending any reasonable amount of traffic to a website page, it’s a good move to split test the page. It’s far easier with a WordPress website.
(1) OptimizePress Experiments.
I LOVE OptimizePress. It’s a page builder that installs in WordPress and allows you to set up quality pages with opt-in forms really fast. It also has an Experiments function for A/B testing.
(2) Visual Website optimiser ( VWO )
VWO is an online service that will work on any website as far as I’m aware. I’ve used it a lot. It imports a page from you site and allows you to modify it then sends the traffic to both variations. You can also do far more complicate experiments and their lowest package isn’t expensive.
(3) Google Experiments
These are a neat way to go about it which works with any type of website. You set up 2 or more pages to run simultaneously and divide the visitors between them. Then Google shows you the results, like the picture on the next page …
Here are the steps to set up a Google experiment.
(3) Simple A/B test WordPress plugins.
If your website is built with WordPress there are a number of simple A/B test plugins that install very easily and cost nothing or very little … https://wordpress.org/plugins/tags/split-testing/
There are a lot of other services that help with online experiments – Unbounce & Optimizely are a couple more – but for most businesses, a simple solution will improve your website’s marketing performance without too steep a learning curve.