I Want To Split Test Headlines For My Squeeze Page

[simmonsmike7]

I Want To Split Test Headlines For My Squeeze Page

...But I don't know what software to use, or how it's done?

I'm a newbie at split testing, but I can figure it out if you tell me what to use to do it - at the moment I just want to test headlines.

Please advise Warriors!

[AdInventive]

PHP Code:

<?php
if(rand(0,1) == 0) {
header("Location: http://www.website.com/page1.html");
} else {
header("Location: http://www.website.com/page2.html");
}
?>

Make that your index.php page or whatever other page traffic is going to.

Really simple way of doing it. There's probably better ways out there but it's always worked fine for me.

[simmonsmike7]

Alright umm where would I track these results?

Or would the opt-in forms on both pages by different?

Please explain... thanks.

[AdInventive]

Just make an exact copy of whatever your page is now. On the copy you made change whatever you would like to test. Put both of the URL's in the code above.

Use whatever tracking software you normally use. Presumably you use aweber, I am not too familiar with it but I would assume they have some sort of way for you to track your subscribes from different sources.

[Pete223]

Originally Posted by simmonsmike7

Alright umm where would I track these results?

Or would the opt-in forms on both pages by different?

Please explain... thanks.

Go back in your Aweber web forms generator, click copy the form you are using
for your initial page and add a tracking code where it says Ad Tracking and save.

Copy/paste the code on your web page you want to split test.

Both forms will add subscribers to the same list but you'll know from which
page they're coming from!

Kindest Regards,
Pete

[Dan Grossman]

Use Google Website Optimizer. Free split testing tool.

[Scott Stamper]

Originally Posted by Dan Grossman

Use Google Website Optimizer. Free split testing tool.

Yes, Google Optimizer is great for split testing. You don't even have to have to have two seperate pages, it will allow you to embed an alternate headline into the same page and rotate the two.

The main problem is that it's not real time. It lags behind a few hours, but otherwise it's quick and easy.

[Joshua Uebergang]

The first suggestions by AdInventive is the simple way to do it, but what about stats? What about conversion rates? Most importantly, what about Taguchi testing? A/B split-testing is cumbersome for most internet marketers because it's SO SLOW. Unless you have tens of thousands of people who can take action on your marketing piece, you must test multiple variables at once.

I used Google Website Optimizer when it came out and I've never used it since. I'll never touch the thing again. Why? Firstly, it takes forever to achieve statistically significant results. Secondly, think what could Google be doing with your data...

Tim over at ConversionChicken.com has a simple solution in most people's price range. His software makes up for the costs in sales you'd lose by A/B split-testing or using Google's software.

[Michael Silvester]

Originally Posted by Joshua Uebergang

it takes forever to achieve statistically significant results

Easy fixed...just get more traffic

You were just spending way too much time down the pub.

Take Care,

Michael Silvester

[Paul Hancox]

Two quick disclaimers. First, I "wrote the book" on split testing. Second, I sell my own split testing software, Power Split Tester - so I may be biased.

That said, there are some things you NEED to know...

(1) There's always a cost involved with free. Google Website Optimizer is good, but it's an "upsell" for AdWords. You need an AdWords account, and so you'll be giving your conversion data to the people who sell you the keywords. Think about it. Plus, any problems and you're mainly on your own as a big company like Google are going to find it difficult to support every free user.

(2) It's NOT going to take you "forever" to split test. Especially since you're testing squeeze page opt-ins, you can get statistically significant results after about 50 or 60 subscribers.

(3) Software that picks a likely winner too early is risky. Let's say Squeeze Page A gets you 10 subscribers, and Squeeze Page B gets you 5... is A better? Not necessarily! The laws of chance are such that it could well be a fluke. You need to get a decent sample, which is why I'm cautious about trusting software that claims to save you money by picking a likely winner too early. You need about 50 or 60 subscribers before you can start to get reliable results.

(4) Make sure your software shows the same version to the same visitor. If the visitor comes back, they need to see the SAME version, not a different rotation. Otherwise, if a visitor first sees Page A, and then comes back and sees Page B and subscribes, which did it... Page A, or B? If you want your split testing data to mean something, make sure each unique visitor sees ONLY ONE VERSION every time.

[TheNightOwl]

I use Power Split Tester and it rocks!

What's it sell for these days, Paul... 50 or 60 bucks?

@OP: You need to be reasonably technically competent, but it's a pretty easy script to install and configure for the most part. I highly recommend it.

On the other hand, if you JUST want a simple split-test for a broad brushstroke comparison of two very different landing pages or offers or, as per your OP, headlines (y'know, all the big stuff), then there are plenty of cheap split-testing scripts around (like, under 20 bucks) that'll do the trick and get you rolling.

Once you've got some money coming in, upgrade to Paul's script.

[Andy]

Hi,

Google Optimizer is an okay choice, I'm not sure I trust their results, but for some reason I've had an awful time running scripts on my hostgator account these days (it didn't used to be like that).

Paul Hancox offers excellent products. While I haven't used his testing script - about the only one I HAVEN'T used (see my hg woes above) - I highly recommend his Small Changes BIG Profits ebook. It's a classic on modern split testing.

His 10% conversion book is a good read also but it has more to do with the sales process than split testing.

Mike Humphries also offers an excellent script. I've used it but, once again, HG has twarted my efforts. Some day I'll fire up the script on another host to verify the problem IS actually the host and not some whacko permissions setting or something else I overlooked.

The difference between Paul's script and Mike's is Mike's also allows you do perform multivariate tests (different elements or "parts" of your page) as well as A/B splits (different pages).

Multivariate testing is quite powerful however you really need a ton of traffic to get effective results you can trust.

For lower traffic testing, a simple A/B script is fine. I've got resale rights to several but I'd think you're better off going with someone who will offer support. As I am not a programmer, I can only offer so much.

BTW Paul's book Small Changes shows you how to test even if you have very little traffic - he calls it "flow testing".

God bless,

Andy

[bestIMtools]

Originally Posted by AdInventive

PHP Code:

<?php
if(rand(0,1) == 0) {
header("Location: http://www.website.com/page1.html");
} else {
header("Location: http://www.website.com/page2.html");
}
?>

Make that your index.php page or whatever other page traffic is going to.

Really simple way of doing it. There's probably better ways out there but it's always worked fine for me.

Just beware that this may significantly affect your search engine rankings. If you make that your index page then there is no content for Google to look at.