Optimizing Your Website for Results


Recently, at the request of our PR manager, I had new headshots taken.    

Not only was my existing headshot a couple years old, but it was unprofessional. It was taken by an amateur photographer in not-so-great lighting.     

So, here are my old and new headshots:   

Dl before after

I'm assuming you can tell that the one on your right is the new one. It was taken by a professional photographer who knew what he was doing.

Now, I'm not writing this essay to talk about branding nor the importance of a good headshot. But, rather, I wanted to talk to you about the importance of split testing.

Let me explain.

Every day, thousands of entrepreneurs visit my website. And many of them come to a page that shows the old picture of me.

So, let me ask you a question: when I change my website to include my new, professional photo, will conversion rates go up?

Will more people download my reports? Will more people purchase my products? Will more people fill out our contact forms? Will more people call Growthink's offices?

Fortunately, we track each of these key metrics. So we'll be able to tell EXACTLY what happens when we change the photo on our website.

So, do you think the new headshot will improve performance?

Well, if you said "yes" I think you're going out on a limb. Likewise, if you said "no" I'm not sure if I'd agree with you.

In fact, the only answer that I like is "who knows."

That's right, while you or I might have a hunch as to the effect the new headshot will have, in actuality, it might help, hurt, or have no effect at all.

We'll have to wait until a statistically significant number of people visit the website, and track the actions they take, in order to figure out the effect.

Importantly, this testing process is known as "split testing." Split testing is a marketing method whereby you compare a "baseline control sample" (my old headshot) to a "new sample in which one variable is changed" (the new headshot) and record the results.

Multivariate testing is the same thing, but when you are able to test multiple variables at the same time (for example if I had one web page that showed my old picture with the name "Dave Lavinsky" in red under it, and a second web page that showed my new picture with the name "David K. Lavinsky" in green beneath it).

Now the key point I want you to understand is this: by continuously using split-testing and multivariate testing on your website, you can improve results and create massive competitive advantage.

So, what should you test on your web pages? Here are a few things:

  • Headlines: the headlines you have at the top of your page

  • Colors: your background and text colors

  • Text Fonts: font sizes and choice of fonts

  • Graphics: the types of images you use (and whether you use images at all)

  • Copy: the text you use throughout your page

  • Positioning: where on your page you include your calls to action (e.g., top right vs. bottom center)>

  • Price: the price at which you are offering your products/services

And when you think about it, the list of things you can test on your web pages is nearly limitless (e.g., adding live chat, adding BBB logos, etc.).

So, here's how it might work. You start with one page on your existing website where you either get most of your traffic, and/or to which you direct specific visitors. This is your "control page."

By installing Google Analytics or some other analytics package on your website, you'll be able to see exactly how this control page is doing. For illustrative purposes, let's say that the goal of your web page is to get visitors to fill out a Contact form, and that currently, 2.3% of visitors complete the form (i.e., your control page has a 2.3% conversion rate).

Then, you start testing variations of your control page (using a tool like Google Website Optimizer makes this easy). You try new headline text. You change the background color. You add new images, etc.

Some of these tests will increase your conversion rate, while others will have no effect or decrease your rate. You find the winners (those that increase conversion rates) and keep testing and testing more variations.

It would not be unlikely for your 2.3% conversion rate to double to 4.6% within a matter of just a few months.

Now, here's where it gets interesting and cool. Because your competitors have NOT split tested and optimized their web pages, they'll still be getting the low 2.3% conversion rate. So, for every thousand visitors, they'd only get 23 leads while you'll be getting 46 leads.

This will allow you to dominate your competitors. You will be able to out-advertise them since your website is so much more effective.

And, they most likely won't be able to "reverse engineer" your web page, since only you'll know the effect that each of the variables you tested had on your conversion rates. They'll be shooting in the dark.

In summary, I want you to think about your website. Whatever results your site is achieving today, they could be MUCH better. Simply by testing key variables, you can find the optimal mix that maximizes your leads and profits. And which leaves your competitors in the dust.


Suggested Resource: The Ultimate Internet Marketing System provides even more instruction on optimizing your website. Click here to learn more.

Share this article:

Get a Free Consultation
with a Growthink Expert

Click Here