Landing page optimization
| Part of a series on |
| Internet marketing |
|---|
| Search engine marketing |
| Display advertising |
| Affiliate marketing |
| Mobile advertising |
Bases for landing page optimization
There are three major types of LPO based on targeting:[1]
- Associative content targeting (also called rule-based optimization or passive targeting). The page content is modified based on information obtained about the visitor's search criteria, geographic information of source traffic, or other known generic parameters that can be used for explicit non-research-based consumer segmentation.
- Predictive content targeting (also called active targeting). The page content is adjusted by correlating any known information about the visitor (e.g., prior purchase behavior, personal demographic information, browsing patterns, etc.) to anticipate (desired) future actions based on predictive analytics.
- Consumer directed targeting (also called social targeting). The page content is created using the relevance of publicly available information through a mechanism based on reviews, ratings, tagging, referrals, etc.
There are two major types of LPO based on experimentation:
- Closed-ended experimentation. Consumers are exposed to several variations of landing pages while their behavior is observed. At the conclusion of the experiment, an optimal page is selected based on the outcome of the experiment.
- Open-ended experimentation. This approach is similar to closed-ended experimentation, except that the experimentation is ongoing, meaning that the landing page is adjusted dynamically as the experiment results change.
Experimentation-based landing page optimization
Experimentation-based LPO can be achieved using A/B testing, multivariate LPO, and total-experience testing. These methodologies are applicable to both closed- and open-ended experimentation.
A/B testing
A/B testing, or A/B split testing, is a method for testing two versions of a webpage: version "A" and version "B". The goal is to test multiple versions of webpages (e.g., home page, product page, FAQ) to determine which version is most appealing/effective. This testing method may also be known as A/B/n split testing; the n denoting more than 2 tests being measured and compared. The data for A/B testing is usually measured via click-through or conversion.[2]
Testing can be conducted sequentially or in parallel. In sequential testing, often the easiest to implement, the various versions of the webpages are made available online for a specified time period. In parallel (split) testing, both versions are made available, and the traffic is divided between the two. The results of sequential split testing can be skewed by differing time periods and traffic patterns in which the different tests are run.
A/B testing has the following advantages:
- Inexpensive because existing resources and tools are used.
- Simple because no complex statistical analysis is required.
A/B testing has the following disadvantages:
- Difficult to control all external factors (e.g., campaigns, search traffic, press releases, seasonality) when using sequential testing.
- Very limited in that reliable conclusions cannot be drawn for pages that contain multiple elements that vary in each version.
Multivariate landing page optimization
Multivariate landing page optimization (MVLPO) accounts for multiple variations of visual elements (e.g., graphics, text) on a page. For example, a given page may have k choices for the title, m choices for the featured image or graphic, and n choices for the company logo. This example yields k×m×n landing page configurations.
Significant improvements can be seen through testing different copy text, form layouts, landing page images and background colours. However, not all elements produce the same improvements in conversions, and by looking at the results from different tests, it is possible to identify the elements that consistently tend to produce the greatest increase in conversions.
The first application of an experimental design for MVLPO was performed by Moskowitz Jacobs Inc. in 1998 as a simulation/demonstration project for Lego. MVLPO did not become a mainstream approach until 2003 or 2004.
MVLPO has the following advantages:
- Provides a reliable, scientifically based approach for understanding customers' preferences and optimizing their experience.
- Has evolved to be an easy-to-use approach in which not much IT involvement is required. In many cases, a few lines of JavaScript allow remote vendor servers to control changes, collect data, and analyze the results.
- Provides a foundation for open-ended experimentation.
MVLPO has the following disadvantages:
- As with any quantitative consumer research, there is a danger of GIGO (garbage in, garbage out). Ideas that are sourced from known customer touchpoints or strategic business objectives are needed to obtain optimal results.
- Focuses on optimizing one page at a time. Website experiences for most sites involve multiple pages, which are typically complex. For an e-commerce website, it is typical for a successful purchase to involve between twelve and eighteen pages; for a support site, even more pages are often required.
Total-experience testing
Total-experience testing, or experience testing, is a type of experiment-based testing in which the entire website experience of the visitor is examined using technical capabilities of the website platform (e.g., ATG, Blue Martini Software, etc.). Rather than creating multiple websites, total-experience testing uses the website platform to create several persistent experiences, and monitors which one is preferred by the customers.
An advantage of total-experience testing is that it reflects the customer's total website experience, not just the experience with a single page. Two disadvantages are that total-experience testing requires a website platform that supports experience testing, and it takes longer to obtain results than A/B testing and MVLPO.
See also
References
- ↑ Alex Gofman, Howard Moskowitz, and Tonis Mets. 2009. Integrating Science into Web Design: Consumer Driven Website Optimization. The Journal of Consumer Marketing, 26(4): 286-298. doi:10.1108/07363760910965882.
- ↑ Matthew Roche (2005-12-19). "Landing Page Testing Best Practices". Site is Dead. Retrieved 2007-07-02.