How do you get peak performance out of a marketing piece like an ad or web page?

Three words: test, test, test.

You test your marketing materials by running different versions against each other to see which one performs best. Some of your audience see one version, some see another, and you record which version they respond to the most.

Once you know that, you keep the best-performing version and run it against other versions until you find one that performs even better.

Imagine — every new improvement you add builds on the proven winners from before! You’re creating an upward spiral of improvements that can send your results through the roof.

People commonly find that when they’re just starting out, their conversion rates are dismal — as low 0.25%. But it’s very common once they start testing they see their conversion rates leap with every new test!

There are different testing methods:

A/B split test

This is the simplest and most popular method. In this type of test you create two different versions of, say, a web page. You direct people to each page randomly and track your results to see which page works generates more leads, conversions, opt-ins, or whatever else you’re aiming for.

Split testing can be done by running two entirely different versions against each other. But if you do that, you won’t be able to tell which individual elements are responsible for the better performance.

Say your website sales go up after you change your headline and images and add more benefits — you don’t know if some changes improved your results and other changes actually hurt them.

It’s far more powerful to test one version against a mirror version that has just one two variables changed. That way you always know exactly which variable is working better.

Multivariate testing

Multivariate testing tests many different elements of your web page at the same time to figure out what the winning combination would be for the best results.

So for instance on a web page, you could test two different versions of your:

• call to action (C1 and C2)
• top-fold graphics (G1 and G2)
• headline (H1 and H2)

With the help of software, your traffic gets split among the different combinations of these variables.

So someone visiting one particular page of your website may see C1, G1 and H2 but does not buy. Someone else sees C1, G2 and H2 and makes a purchase. You can imagine all the different variations that can be made.

You need to test multivariate testing with a fair amount of traffic to find the winning combination. The more elements you you want to test, the more page combinations you’ll create, so the more traffic you need to make this work properly.

What variables should I test?

Web page	Pay-per-click ad
Header Headlines Subheads Images Formatting Navigation Font, size, colors Opt-in offer Opt-in box Opening paragraph Benefits Testimonials Body text Anchor text for links Copy length Guarantees Bonus offers Call to action P.S. Price points Order page Thank-you page	Headline Ad copy Keywords Capitalization Bonus offer Call to action Price/no price Ad/landing page combination Ad format

Starting your first split test

First, decide what results you’re tracking. Do you want more traffic, more opt-ins, or more conversions, for instance?

Your original page is called the control page. This is the page you want to measure results against to see which variables are working and which are not.

Now if you’re doing A/B split testing — which is a great place to start for beginners – remember to change only one thing at a time or you’ll have no idea what is working and what is not!

With the multivariate testing, it’s quite different. You’ll be testing different combinations against each other, which means you can test out more variables in a shorter amount of time at a smaller cost.

Every time you do a test, you get to gather more information about your users and find out more about what your audience is after. This not only helps your conversion rates, it can lead to further product development!

I do caution that the biggest mistake people make when they start testing is not getting enough traffic to yield solid results. There is not a specific number I can give you but the more people you test, the accurate your results are going to be.

There are a lot of tools out there for testing web pages, but Google has a free one called Google Optimizer that is very good and will let you do all three kinds of testing.

Of course, this kind of testing is not just for your web pages. There are tools that help you test your PPC ads too.

In all the pay-per-click programs (Google AdWords, Yahoo Search Marketing, and Microsoft adCenter) you can create different ads for the same ad group. Then you can test out different benefits in your ad, different offers like free report or shipping or bonus gift, and see which one gets the best click-through rate.

Remember you’re paying good money for every click-through. Why not make the most out of it and really improve the results of your bottom line?

Being able to test every element of your web page or PPC ad is what puts Internet marketing head and shoulders above all the other types of marketing… so take advantage of it!

OR

Let's our Team Experts to Fix your website problem for you for Just $2.95

"Do Your Website Sales SUCK? Hire A Team Of Experts To Help Fix The Problem For Just $2.95!" This crackshot team of Internet wealth experts has already generated over $60,000,000 in online sales. Now, for just $2.95, they'll spend the next 30 days working with you to: >> Figure out where you website is leaking money, >> Make simple changes that can increase sales by 300% or more, >> Blast your website to #1 in Google, ... plus tons more. To get your website reviewed for just $2.95 click here now.