Here I descibe in detail my approach to taking a working control landing page and turning it into something all together more effective. My landing page optimization process, fully at work. The current â€œcontrolâ€ page The current landing page has evolved through vigorous AB split testing on PPC. We utilize our highest volume term for this testing, the â€œLoansâ€ exact match keyword. We have very similar variations of this page for other terms â€“ with the only difference being the header â€œSearch 100s of loans, unrivalled access to 100s of loans, exclusively for homeownersâ€. This text exactly matches PPC display copy we use on the appropriate terms. We also use this page on a variety of banner and email campaigns â€“ where this header section changes to match the message and graphics of the source media. Everything else on the page remains the same. This landing page is a three page application form, for the purposes of lead generation. We have tested single page and two page applications but neither beat this control. Motivation Our research has shown that approximately 80% of our customers take out a loan for debt consolidation. Our domain name contributes to this, yet we do not currently target our copy specifically at this market. The current thinking is that we can offer loans for any purpose and by targeting consolidation we will lose customers. We have attempted to address this by tailoring our core value proposition to a consolidation audience. Also by using a page with more copy, we hope to be able to address the concerns and meet the expectations of customers coming from different sources e.g. PPC customers are actively searching for a loan while banner customers were not. The optimized page is designed for use across PPC â€“ â€œloansâ€ keyword but also more specific â€œsecured loanâ€ and â€œdebt consolidationâ€ group terms. We will lead new PPC campaigns with a debt consolidation message â€“ similar to the lead message on the new page design. Our current page requires three pages be completed. Our data is automatically processed and routed to different parts of our business depending on how the form is completed. We cannot currently change the questions we have to ask. We believe that users arriving on the first page are expecting an aggregator style loan search â€“ an expectation which we cannot fulfill. We also believe that requiring a user to stick with a single process over three pages is asking a lot. On the optimized page we have drastically changed the â€œangleâ€ that our app form takes. We have turned the first page into a detailed â€œsavings calculatorâ€ targeted specifically at consolidation customers. By asking the certain questions up-front and including â€œcurrent monthly repaymentâ€ and â€œcredit ratingâ€ fields we are able to calculate how much money we can save the customer on a monthly basis. We have also included the email field (optional) at this stage. If they provide their email address then we will email them their saving quote â€“ plus we can send a customer service email if they fail to complete page two. As our processing requires an email address for a full application we will need to ask for the email (mandatory) on page two if the user does not already supply it. We are trying to turn a process which spans three pages look like two separate processes, each of which are shorter and provide feedback to the customer. Value Proposition As our product (secured loans) is a commodity which differs very little from company to company â€“ and especially as we are a broker offering exactly the same loans as other brokers â€“ we have struggled to find a compelling value proposition. Our company has built itself on excellent customer service, highly efficient marketing and processing, which we realize are not unique enough to form a value proposition. Our current control uses our PPC copy as it”s value proposition â€œSearch 100s of loansâ€ which a number of our competitors also use in one form or another. We do not have the most loans, so we have decided to move away from this message. We have tried to differentiate between the first and second pages of the application in order to compel customers to disassociate it with a single, long form to fill-in. Our first page leads with the value propositionâ€¦ See instantly how much money you could save every â€“ FREE. This fits the new style app form, which will calculate the customer”s savings and display the result on the second page. The remaining copy on the page all tries to reinforce this message and explain what the user will get when they complete the form (this page). We have also tried to emphasize the fact that this is a totally none-committal stage in the process. The second page will calculate the customers saving and display an appropriate message (we will have a generic message if we cannot save them money e.g. the user fills in zero for their current monthly repayments)â€¦ We could save you Â£XXX every month. The second page leads with a saving we have calculated from the data the customer has provided. We hope this is a strong message to get them to apply for a further quote â€“ something we try and emphasize in the remaining copy on this page. Friction Several major areas of friction were identified on the current page. Number of questions Unfortunately this cannot be changed as these questions have been integrated into a rules engine for prioritizing and segregating our leads. Hopefully this has been alleviated on the optimized page by splitting the process into two separate forms â€“ each of which has a clearly identified result for the customer. Difficulty of questions Some of the questions have relatively easy to misunderstand answers â€“ such as are you a homeowner: yes, no, yes â€“ but I don”t pay a mortgage. We have set default values where appropriate on the form in the optimized page but we cannot change these questions. Eye flow The eye flows from the logo (which is oriented inappropriately as it has been taken from offline designs) and across to the value proposition. This initial step isn”t too bad but the eye then proceeds down to the application form. This is an obvious problem because as yet we have told them virtually nothing about who we are, what we offer and why they should deal with us. The optimized page simplifies the layout significantly with just a single column. We have also adopted a much less cold approach to communicating our messages and have added lines to the application form to help explain and reassure the customer. Three page process The first page serves no purpose other than to collect data â€“ and it there is no logical text to place on the button because of this. When the user sees â€œNextâ€ they must be thinking â€œhow many more pagesâ€. We had previously tried to solve this by placing all of the questions on a single page â€“ but we believe this made matters worse because the form became daunting and intimidating. For the optimized page we have broken the process down into two distinct and useful stages for the customer. The first page now has a distinct purpose and a result on page two. Page two then becomes the â€œapplicationâ€ in the customer”s mind (although really we need data from both pages). Anxiety Several major areas of anxiety were identified on the current page. What am I committing to when I complete this form? Splitting the process into two distinct phases should help this as the first stage can be totally none committal. We mention this several times in the copy and the way the form is presented should convey to the customer that their result will be displayed on the second page but that this is not an application, or an application for a quote. Page two of our process is an application for a quote â€“ still none committal but we will contact them. Our current control does not explain this at all and is presented as an â€œapplication formâ€. To me that suggests a certain level of commitment, while a â€œquoteâ€ does not. This stage of the process is simply a quote â€“ so our new page now presents it as such. Nature of the questions We ask for a lot of personal information, especially a lot of contact information. We cannot change this but have tried to explain how and why we need the information â€“ justifying it is much as possible. We have also moved our VeriSign seal from the bottom of the page to adjacent to the form buttons. Incentive We have tested various incentives on our direct mail campaigns with varied success. Given the nature of our business we can only offer incentives to people who actually take out a loan with us, and as such it is a decision of some gravity. We have decided not to add an incentive at this stage but it is something we need to pursue in the future.