Though a test may improve a page’s performance, its greatest gain often comes from what it teaches you about your visitors. This is because the value of customer insights extends far beyond a single test. However, often the potential insights of a test are overlooked, and consequently, lots of revenue is being left on the table. So, in our October 20th Web clinic, Dr. Flint McGlaughlin provided a robust methodology for proper test interpretation to ensure the greatest long-term return on marketing investments.
In this Web clinic, we studied a live experiment from our research laboratory and conducted a hands-on, workshop-like analysis of the results, revealing the insights we discovered from the experiment. We walked-through this case study – which produced a 23% increase in revenue-per-visitor – from the unique vantage-point of one of our own research analysts.
Dr. McGlaughlin then taught the three key questions that must be asked of every test in order to maximize the insights gained:
QUESTION: What does this test say about my visitors?
- Proper interpretation reveals answers to questions like: What motivates my customers? What do my customers value? What causes them the most anxiety? Where are they in the conversation? These are just some of the potential insights discussed in this section of the web clinic.
- We also deal with the question, “Can reliable insights be drawn from a test that compares multiple variables (variable cluster)?” Our very own Director of Sciences, Bob Kemper, provided the scientific rationale behind running tests with “variable clusters”.
QUESTION: Where else might these insights apply?
- The greatest impact gained from a test is actualized as test insights are applied to other marketing campaigns. In this section of the web clinic, we discussed the importance of identifying the two types of insights gained from any test and the subsequent application.
- Context-Specific Insights – Insights that are applicable to scenarios similar to that of the test environment.
- Meta-Theoretical Insights – Insights that contribute to a deeper understanding of the cognitive psychology of human nature.
QUESTION: What do these results imply we should test next?
- This section of the web clinic discussed how future tests will be directly dependent upon the degree of precision with which you can identify the elements that generate the greatest favorable variation. And that testing should progress from variable clusters toward single-variable experiments.
- Returning to the experiment analyzed throughout the entire web clinic, we revealed a follow-up test our team is running in the coming weeks. We engaged the audience for their feedback on the upcoming experiment.
Finally, Dr. McGlaughlin and the MECLABS Conversion Group team conducted the first-ever Test Strategy Recommendation Session of audience-submitted campaigns. Looking at the test history of audience-submitted online Web pages, we recommend future test strategies based on a decade of optimization research.
You can view the clinic replay, or listen to the audio recording (mp3), to learn how to begin gaining more insight (and ultimately more return on investment) from your own tests.
Download the MarketingExperiments Quarterly Research Journal:
View a replay of this presentation:
Presentation will open in new window
Credits:
Presenters
Flint McGlaughlin
Bob Kemper
Nathan Thompson
Writers
Austin McCraw
Daniel Burstein
Bob Kemper
Additional Contributors
Flint McGlaughlin
Nathan Thompson
MECLABS Conversion Group