I have spent the past two years researching and testing what works on the web. I have studied the 20th Century’s great marketers, applied best practice to dozens of web sites, and collated the evidence.
The most valuable thing I have learned is that I know absolutely nothing about what makes people buy.
Really. The number one insight I can give you is that whatever you think is likely to work probably won’t.
Consider this as Rule #1. There really are no other rules that matter.
One reason for this is that your customers aren’t you. They think differently, and they have a very different perception of what they need and what you offer. It’s like there’s a black box that you cannot ever fully understand - the psychology of your market.
So what can you do? How can you sell when you don’t really know what people want?
There are two things you have to do, and one thing you must never do.
The thing you should never do is to assume you do know. Your first best guess is almost certainly far from correct.
What you should do is: test! Previous generations of marketers knew this, and practiced it religiously. They split-test print ads and mail-outs because they knew that it would generate better returns. It is no different today.
The good news is, the web lets you do that more easily and quickly than ever. Where it would have cost thousands of dollars and weeks of time to split-test an advertisement in a newspaper, you can test endless variations of your messaging on your own web sites very rapidly and for very little cost.
Today, my company’s core revenue comes from subscriptions to my Pro Web Design Course. I recently spent days crafting a new landing page to promote the course (http://www.prowebdesigncourse.com/). But Rule #1 reminds me that my first best guess is not right. So I’m testing.
See the image below (you may need to “show images”):
When I created this page, I also created a few variations of the main headings and introductory text. I then used one of the powerful testing tools on the market to serve up random combinations of this content to all visitors.
The tool (Visual Website Optimizer) tracks how many people who see each content combination click through to the pricing page.
This image shows that 20.5% of people who see my first best guess (”Control”) are persuaded to click through.
But some of the other combinations engage 42% more visitors. That’s potentially 42% more income for my business - from a test that took me less than an hour to set up!
The results are not yet conclusive, but when it has finished I will know which heading and intro text works best.
You know what I’ll do then? I’ll test again. I’ll see whether I can come up with alternative content, rearranged content, more content, which will compel even more people to click through.
Plus we’ll also be testing the pricing page itself, to see how many more people we can persuade to take the final step and sign up.
And that’s how it goes. I don’t know anything. I may have some smart guesses, and some of them may work better than my previous guesses. But until I test, I won’t know for sure, and maybe I’ll never find out what works best of all.
Consider the economics for your business. Setting up A/B or multivariate tests with a tool like Visual Website Optimizer takes no time at all, and yet can potentially double your revenues, double your leads, double whatever you want people to do on your web site. Or better! I know this, because I’ve done it.
You don’t know what works. Until you test, you’ll never be sure. How much money are you leaving on the table? Want to find out?
Best wishes,
Ben
Ben Hunt Limited
50 Penrhyn Road
Sheffield, UK