Picture the scenario; your marketing people have written a new slogan, your web designer has made a new layout and your graphics guys have designed a new logo. Your MD likes the logo but hates the copy, your sales Director loves the copy but hates the layout and you love the layout but hate the logo.

You could argue these points for ever, you could ask more people in your organisation, you could employ consultants to decide for you, you could flip a coin or ask Paul the psychic octopus.

And you would probably still get it wrong.

The best, simplest and cheapest way to make the decision is to test. In the scenario above, you have two logos, two layouts and two slogans. If you were to make every possible combination you have eight different versions of the page – 2 x 2 x 2 = 8.

Multivariate testing lets you test each version of the page on real life customers – without them even knowing you’re testing.

Multivariate testing generates all the possible versions of your page, displays them randomly to each visitor and measures the conversion rate for each one. Not only that, it tells you which area – logo, slogan or layout – has the biggest effect on conversion.

You get to make the decision based on real results from your visitors, not on the personal opinions and vested interests of individuals. Clever huh?