WiderFunnel Case Study
Landing Page Optimization of Newsletter Permissions Improves Opt-in Rate by 12%
The Client: Tourism British Columbia
As a marketing-oriented organization, Tourism British Columbia’s mandate is to promote the growth and development of a $9.8 billion tourism industry through innovative programs and industry development initiatives. Its consumer-facing website, HelloBC.com, is the official travel planning site for the Province of British Columbia, Canada, providing detailed visitor information for over 130 communities as well as 3,000 approved accommodations. It is a popular site for local and international travelers looking for information on places to go and things to see in BC.
The Importance of Permission
One of Tourism BC’s key business goals is to grow its marketing database leading up to the 2010 Winter Olympics in Vancouver, BC, and their campaign Landing Pages play a starring role in making this happen.
Tourism BC’s conversion-optimized Landing Pages for free BC trip-planning Guides have become very successful at converting web Visitors into qualified Inquirers, but the ongoing value of these information Inquirers is lost if they do not opt-in to receive further communications.
Tourism BC recognized the importance of opt-in permission and knew they could do more to increase the opt-in permission Conversion Rate.
Conversion Rate Optimization for Opt-In Permission
As part of WiderFunnel Marketing’s ongoing conversion optimization work with Tourism BC, WiderFunnel planned, designed and implemented an A/B/n content experiment to increase the permission opt-in rate on the Landing Pages. See Fig. 1 for the original ‘BC Escapes 2008’ Landing Page and the location (outlined in red) where WiderFunnel placed a variable content SwapBox for testing.
The goal for Tourism BC is, first, to increase the conversion rate and, second, to understand which specific factor had the most influence. The experiments are designed, beginning with the hypothesis creation, to increase the conversion rates rather than as research projects.
The Landing Page Optimization Process
Step #1: Hypothesis Creation
One of the most important steps in WiderFunnel’s Funnel Experiment™ (or FunEx™) process is hypothesis development. No experiment should ever be run without a valid hypothesis that will be supported or discredited by the experiment.
The original permission box copy on the Landing Page, shown in Fig. 2, had not been modified for several years and Tourism BC knew it needed to be re-evaluated from a Conversion Rate Optimization perspective.
Tourism BC worked with WiderFunnel to develop three hypotheses that would be tested in the SwapBox area. Each hypothesis would then be used to create three alternative presentations of the SwapBox’s content.
The three hypotheses were:
- A design and copy variation with a popup eNewsletter preview and more benefit-oriented copy will improve clarity and increase the opt-in conversion rate
- Reducing the number of opt-in choices from three to two will increase the opt-in conversion rate
- Moving the permissions boxes higher in the form, below the name and address information, will increase the opt-in conversion rate
Step #2. Develop Alternative Design and Copy
Once the hypotheses were agreed-upon, WiderFunnel used them as the instructions to design the layout and copy for the each of the three SwapBox variations.
The alternative designs included a prominent preview image of a typical Tourism BC newsletter and benefit-oriented copy. Both, the design and copy changes, were intended to improve the Clarity of the presentation. “Clarity is one of the six Conversion Influence Factors we know affect Conversion Rates,” says Chris Goward, Co-Founder of WiderFunnel, “Most marketers can dramatically improve results just by spending a few rounds focused on improving the clarity of their presentation.”
Step #3. Run online page tests
Using Google Website Optimizer as the testing platform, WiderFunnel ran the experiment to determine which of the four variations (including the original ‘Control’ version) would produce the highest opt-in permission rate.
Importantly, the experiment was set up to track a goal only when a visitor completed the form and had one of Tourism BC’s permission boxes checked.
In this case, due to web traffic levels, the first test was planned as a variable cluster, meaning that several conversion hypotheses were included in each variation. Further testing will be required to determine whether the new copy or the preview thumbnail, for example, were the most influential factors. See Table 1 for the hypotheses included in each test variation.
Results: 12% more opt-in database subscribers
The experiment completed with statistically significant results showing that WiderFunnel’s new design with improved copy and the eNewsletter preview generated a 12% lift in the opt-in conversion rate.
This combination of benefit-oriented copy and a Newsletter preview image makes a significant improvement in persuading visitors to subscribe.
The second and third hypotheses, which included reducing the number of opt-in choices and moving the opt-in fields higher on the form, did not significantly impact conversions.
More testing will be required to determine whether the copy or image played a more important role in the conversion rate lift and whether 2 or 3 permission boxes have any effect.
Tourism BC has now implemented this winning permission box presentation onto all Landing Pages and benefits from a larger database of prospective travelers generated from the same demand-creation marketing budget.
Continuous Improvement Philosophy
Tourism British Columbia is an innovative marketing organization and a leader in using integrated marketing strategies to promote the province internationally. The organization’s work with WiderFunnel to optimize its Landing Page and website conversion rates is one example of its commitment to continuous improvement.
Learn more about the services mentioned in this case study:
- Website Conversion Optimization
- Landing Page Optimization
- The Kaizen Method
- A/B/n Split Testing
- Multivariate Testing