ocean was going to be bloody red. We decided to use a product called Unbounce to quickly build out variations of the value proposition and test them by using targeted online advertising campaigns through Google and Facebook. This effectively put the current site on hold. We wouldn’t disrupt any of the complex backend. The developers wouldn’t have to waste any time, energy, or money on fixing it for the experiment. Instead, if any suspects did go through our Landing Page experiment, we would concierge them (as Jared did, as described in Chapter 7) through the process of booking a treatment center. Figure 9-17 shows what the test landing page looked like. To drive traffic to our landing page, we did targeted advertising on Facebook. We ran ads to users who lived in wealthy neighborhoods in Los Angeles. Here is where you want to be careful, because it’s easy to spend more money than you want or need to for a small trial. That’s why we came up with the following boundaries: We decided on an initial budget of $500. We determined the key words for the campaign. We finalized the advertisement copy to ensure that the messaging would be strong enough to drive suspects to the landing page. We established the customer demographics (such as education, city, and age). We decided on when exactly the campaign should start. We placed the campaign, which you can see in Figure 9-18, and then we waited!