A/B Testing: 5 Things You Need to Do before you start
Since they are quite a buzzword these days, you also must have heard something about A/B testing. If not, check them out as soon as possible, as truth be told, they can be extremely helpful at improving conversion on your website. However, they are not eligible at just anytime, they are actually only helpful after reaching a certain stage of startup development. Unfortunately, most websites are not yet at the point where they can start A/B tests and will most probably never get there.
So, You! You’d better start thinking about improving your conversion after validating your startup idea and reaching the point of product/market fit. The naked truth, as I mentioned, is that if you’re still struggling to acquire your first users, A/B Tests won’t help you much. You are better off focusing on identifying the needs of your target group and improving your solutions. Keep in mind that in the world of consumption, endless choices, and possibilities, no one will sign up for a useless product simply because it has a fancy pitch or catchy colors. So remember: A/B Tests can boost your performance at engaging users and customers but you need a certain amount of traffic to benefit from them.
At the outset, you must opt for analytic software. There’s a lot of those to choose from, but I would advise you to choose from the popular ones: Unbounce.com, UsabilityTools.com, Optimize.ly, or VisualWebsiteOptimizer.com. Don’t take too long to think which one to single out, they’re all brilliant. And forget about building dedicated test software. It’s like developing your own Gmail: exciting but stupid.
Here’s what you should keep in mind before you start A/B testing:
Do not take anything for granted.
It’s admirable that you read case studies and learn about working A/B Test practices, but bear in mind that you must also search for an improvement of your own. Reading what pushed 37signals to rock their conversion will definitely give you some perspective and help a lot. However, you need to test what works in your very own case. Every product is different and so is its audience.
Test your Software.
It might sound laughable but it is pivotal – make sure your test software is accurate. A short A/A test will help to find out whether the software is cheating on you when it comes to the results or not.
Do not get bored.
A/B Testing might never be over, there always is room for further improvement. Each and every tiny detail, such as red gradient or its absence somewhere out there, on your website, even on the minor button can have a major impact on an overall conversion. So don’t underestimate the influence of any change until you test it.
Test whatever comes to your mind.
Don’t stop at CTA, pitch, layout or colors. Take your time and examine the influence of mentioning more/fewer unique selling points, changing the price, or maybe even implementing a completely different pitch on your landing page. You would be surprised what a result it can bring.
Get into the habit.
Create a spreadsheet and write down all the ideas that pop up into your mind. Be prepare to test various solutions and execute different changes for months. Remember not to give up on any idea. Each test should be conducted for one to two weeks and lead to no less than 100 conversions. Only at that point can you be sure that no random mistake appeared. For better clarity, you can even create a tests table or a calendar. Do whatever helps you to keep on optimizing your website.