Is your pay per click or other advertising campaign as effective as it could be?
Chances are, that it could always use some improvement. No one is at 100% conversion rate online so there is always room for opportunity to improve, but only if you can identify that opportunity and act upon it. That is where split testing comes in.
Read on to find out exactly what split testing is, and how it can help your marketing efforts.
What Is Split Testing?
Split Testing refers to trying out more than one type of promotion at the same time. It might be that you run two types of advertisements at once. It could also be that you send out a pair of similar email newsletters, each to half of your database.
The same principal applies to a sales letter, a printed piece, your website landing pages or anything you use for marketing.
Why Bother Doing Twice The Work?
With just one banner ad, how you can tell what is really inducing someone to click on it? Is it the colour, or the animation, or the text, or the offer? There is no way to tell until you start isolating variables.
Say for instance you’ve created an advertisement for your pay per click campaign based on the best keywords and content that will induce web surfers to click on it. You may think it is successful. Perhaps you’ve seen an increase in traffic and your sales have jumped. Obviously the ad is doing what you wanted it to do.
Creating another ad that is very similar (a mirror ad) but moves the text around a bit may give you further information. If your current banner includes an offer for a 10% discount along the bottom of the box, create another banner that puts that offer at the top. This way you can see which one is more effective, and determine if the 10% discount, or another feature, is driving traffic and sales.
How To Perform Split Testing
In a pay per click campaign, you will want to run each of your ads for at least 30 clicks. Keep a spreadsheet that shows the click through rate you get for each one. If the second banner is more effective than the first, determine the change that made it so. Then take it a step further. Take the second ad, and tweak it again.
In the above example, perhaps you have determined that the prominence of the 10% discount is the feature driving traffic to your site. Your next ad running simultaneously could include the price of your most popular product. Make another comparison and then another ad.
The longer you keep up this process, the more you will be keyed in to the features that best attract customers. This is extremely valuable information that can be used in all your future marketing efforts.
Of course, Split Testing can be performed on your website, too. Try changing the headline, sub-headline, and first paragraph of your home page. Use the two different landing page URLs on each of two concurrent advertisements and see which one generates more conversions.
When it comes to marketing, too often our efforts are left up to hunches or supposition. Why not go about it in a more scientific manner and analyse what is really driving your sales?
Split Testing will set your promotions on the path to sales success.
Have a most outstanding day.
Sean RasmussenAussie Internet Marketing
www.SeanSEO.com © 2008 - 2010
{ 10 comments… read them below or add one }
Thanks, I have been looking into split testing lately myself as I needed easy to use software for my company. After searching the web I reviewed several software packages (all within a certain price range) and eventually decided to opt for Logaholic Web Analytics. Split testing is one of many features in the Logaholic software and it does everything I need, fast and simple. It helped me a lot identifying conversion rates to each of my target pages.
Hi Sal,
Great to hear you have found something that is working for you. Split testing is imperative if you want to get the most out of your marketing strategy.
Regards – Sean
To me, split testing seems like a logical process to work through to improve conversion rates. The key is just not changing too many variables at once. What would also be key is to keep good documentation of the changes as you move further through the process.
Of course that could be overload for some people and having some software like you’ve mentioned Sal sounds like a great way to keep track and keep it simple.
With internet marketing and the vast amount of tools to help, simple and easy to follow is good [for me].
This will be great for when (if) I get into PPC
Hi Sean,
Thanks for the great explanation of split-testing and how I can use it.
I find I get a lot of ideas by using the local free community notice board. I try my ads out there first and see what the response I get. It’s funny, this helps me not only to find out what works before placing an ad online, but the mind seems to create new and sometimes better options while the ad is sitting there being viewed by so many people.
Great strategy Wal, thanks for sharing it here!
Split testing ads in ppc really does improve your conversion rate. What I think is a good ad often is not the one that works. I started out by Googling the keywords, seeing what the highest ranked competitors had written and tweaked them around. Another way was to take a short interesting headline out of the landing page and use that, which really did work well.
Hi Sean,
thanks heaps for this detailed explanation on how we can be using split testing to improve our conversion rates. I have heard it often, but never seen such a detailed explanation of how to actually work it. Tks.
I was never quite sure how one went about using split testing to improve conversion rates. This article gives me a much better perspective on what I have to do to carry out a split test. It’s interesting to note that something as simple as the positioning of a line of text, the 10% discount offer for example, could have an impact on conversion rates.