An Open Invitation: Conversions don't start on the landing page

Some old friends found my blog and asked me a few novice questions personally. I realized that I have not talked much about traffic monetization strategies. There are excellent blogs out there that give solid advice on how to make money online. As has been my practice so far, I will not water what is already wet, but this is a topic that needs some attention. 

There is one fundamental thing I want my readers to understand. While there are many (often conflicting) opinions out there on the Web about making money—about what works and what won’t, what sells and what doesn't, which merchants or traffic sources are better—the only way to know for sure is to test it yourself.  

This has always been my way of doing things. Any new idea that comes to mind or that I learn from someone else, as long as it the basic logic test (“Does it make sense?”), I try it out. I love testing, so I am certain you won't be surprised to learn that I especially love split testing. Imagine how ecstatic I was when Google decided to release the Google Website Optimizer (GWO) for free to Adwords advertisers. Offermatica charges $10k dollars a month for a similar product. Why? Because split testing works! That is where the money is. 

Let me explain briefly how the GWO works (you can learn more from their tutorial) before moving on to what I would like it to do—my dream testing machine. The idea behind GWO is that you select multiple landing pages to compare. Each page must have a persuasive message and a call to action, which is ultimately what you want to test. You can define sections on your page with variations in the heading, an image/offer, or a button. You put some JavaScript code in your landing page that defines the regions, and some tracking code in the next page that defines the conversion. In GWO you create the content for each variation you defined and the amount of traffic you want to send to the test pages. 

GWO will randomly present visitors different combinations of the landing pages. The interesting aspect of split testing systems is that they use advanced statistics so that they don't need to test every possible combination, but only the statistically meaningful ones. (I am not certain this is the way GWO works, but I know that Offermatica's product does it this way.) After a few days, you will be able to see detailed reports that pinpoint the best-converting landing pages. You make those your default ones, and let the money flow in! 

It is important to understand the wisdom of letting these experiments run as long as possible. I've seen cases where the exact same experiment run for a couple of weeks in different periods yields completely different results. But there is another point to consider here. When somebody reaches your landing page, they already have a precondition of what to expect. They followed a link, a text ad, a banner, a review, etc. This “invitation” has an intrinsic promise that the landing page must deliver. Oftentimes visitors will “bounce” back instead of reaching the conversion goal because they were promised something that the landing page just doesn't deliver.  

Google offers a similar split-testing tool for ad variations. The goal is to rotate different ads and identify the ones that yield the best click-through rates. This leads to another insight. Aren’t these two separate tools ignoring an important connection? What if Google linked ad testing with the Google Website Optimizer? Think about your ads as extensions of your landing page, as another section that needs to be tested in the same experiment. Instead of testing the ads only for clicks, Google could help us identify the ads that drive the most conversions. Remember that getting a lot of traffic is good (high click-through rate) but traffic that takes action is even better! 

Every market, every person is different. While we all aim to be successful, we have different goals and expectations. When I get advice online, I know that advice is based mostly on the person's experience. If there is one thing I want you to take from this article it is that you should listen to successful people—but you need to make your own judgments based on your own experiments. 

Let the testing begin!