Sales Page Testing

Change just one thing at a time. Otherwise you won't know exactly what made the conversions better.

Denise

Easy,
There are many things you can test on your sales page or landing page. Believe it or not color can make a big difference in sales.
Also try using different headlines, stories, opt-in box looks.

Of course, test one item at a time. You can give it a timed trial such as 1 week or base it per 100 visitors.
Jeff

Split testing (or A/B testing) is best if you have relatively small traffic volume. Ideally, you'll want at least 300-500 visitors to test as a bare minimum, 1000 is better, but sometimes winners are more readily apparent straight away.

Testing two different headlines, or two different headline colors, for example is easily done with simple split testing. So is running two completely different letters by two completely different writers against each other -- may the best copy win! So is testing two affiliate offers against each other -- may the best product win...

Multivariate testing (testing multiple changes) goes much deeper but each element you add (test) places greater demand on traffic volume to deliver meaningful data. If you're getting lots of traffic, you can literally tweak out every meaningful lift and squeeze out every last penny.

With multivariate testing you can test placement of different page elements, along with attributes of those elements. Say for example, testing an optin box on the top right or mid-screen, orange 180px button or blue 180px button... while you test those headlines mentioned in the split test above, while you test header or no header... and so on.

Testing is the only possible way to TRULY know what works and what sucks.

Very few will ever bother to find out. Isn't that encouraging to know?

What they don't know can kill them and make you rich... : )

Google's website optimizer is free and completely useful. If you've never set up any testing on any of your sites before, it's a great way to learn how to do it.

Brian

It really depends on how much traffic you have, and how many "actions" (i.e. subscriptions or sales) you're getting from that traffic.

Split testing is best if you have relatively low traffic, but the number of "actions" is the most important thing.

I explain this in detail in my book Small Changes: Big Profits, but basically you need to get enough of a sample of actions to get significant results.

You can't just say, Version A got 4 sales and Version B got 6, so B is better... it doesn't work like that.

I recommend at least 50 total actions between the two versions before you can say which is better.

As for what to change, you could change individual parts of the page (such as headline, then body), but I've also done radical changes as well, such as testing two completely different styles of landing page.

The amount of testing you can do ultimately depends on how much traffic you get.

You can use Google Website Optimizer if you don't mind giving Google practically your whole business data.

Thanks Mike for the recommendation.

And I think you're right about Google. I like Google as a company, but with all the data they already have from Search, Analytics and Adwords, I personally don't want them to have my conversion data as well, so they know what profit is being made from each keyword...

... especially when they're charging me for those keywords!

For people with enough traffic, multivariate testing is very useful on a squeeze page, because you can test lots of elements at once, ie. headline, bullets, opt-in form and see which combinations work best.

And in that regard, I'd say Mike's script (Easy Multi Tracking, see his post above) is the best value for money for multivariate testing (without handing your valuable business data over to Google.)

Definitely change one element at a time. Yes, it may be more work, but it is easier to track the results.

For example, testing three or four of your best headlines only, with all other factors being the same, you will have a better idea which headline pulls the best.

Continuing this process throughout your squeeze/landing page will help pinpoint the best combination. By the way, testing the headline is a great place to start.

Hope this helps,

Kelly