A/B testing still dominates the optimization space, particularly for travel brands.
It remains the most used testing method for improving conversion rates.
By contrast, for reasons ranging from testing vendor limitations to practitioner anxiety, multivariate testing (MVT) remains an underused tool.
NB: This is an analysis by Chloë Lind, managing consultant at Maxymiser.
I’m here to make the case for MVT and explain how hotel brands can answer important questions about shifting trends in the travel industry.
For example, decision-making surrounding traveller reviews, one of the most important factors in the online travel experience, can be done with confidence using a well-orchestrated multivariate test.
But first, let’s delve a bit into the powerful voice of the consumer.
It’s nothing new. According to Phocuswright, user reviews rank among the most important features for travel decision-making.
Travel review websites like category leader TripAdvisor are the most common websites/apps used to view traveler-submitted ratings and reviews for the US.
The good news is, hotel brands have the option of displaying a review widget or pulling in TripAdvisor’s reviews on their own sites.
The question then becomes: how can brands use this resource effectively by giving necessary credence to the consumer, without losing their viewers to TripAdvisor?
Furthermore, how can they do so without negatively impacting the brand with less than favorable reviews?
This is when multivariate testing becomes crucial. Rather than just blindly surfacing reviews on their sites, with MVT hotel brands can test to measure the impact on bookings.
A simple A/B test can offer a yes or no answer about displaying the reviews on hotel property pages, and can go further to prescribe what exactly to show, where to show it, and how?
Only an MVT can answer these multi-faceted, interrelated questions efficiently and deliver specific insight into the best way to leverage the voice of the customer.
I’ve designed a test to demonstrate how a MVT strategy shows you the real impact drivers when it comes to the most insightful, granular questions about hosting TripAdvisor reviews on site.
This single multivariate test will answer the following questions:
- Where on the hotel property page (I’ve chosen this page for my example) to show the reviews?
- What specifically are we showing? Do you show three stars and a review snippet? Just stars? Only highly rated reviews?
- How do we show it? Lightbox? Scroll? Link to TripAdvisor?
Because I set up a multivariate test and can track metrics such as engagement, step conversion, booking conversion, and revenue tracking, we not only know the influence of each individual change on every KPI, but we also understand the impact of the actual star rating (1-5) on their user’s behavior.
If it’s not obvious by now, with a multivariate test like this, you can get tons of insight into how to move forward with the inclusion or exclusion of on-page information such as TripAdvisor ratings from a broader business perspective.
On the other hand, if you try to answer complex questions like these with an A/B test, you would need to run four different campaigns, which lacks the efficiency and depth of insight afforded by a MVT.
In order to understand which variant is driving the most impact in a given experience, a multivariate test the best possible approach.
I’ll leave you with one final challenge.
In 2015, make it your New Year’s resolution to execute a multivariate test to answer a business question, big or small.
Once you prove the ROI of your first multivariate test, it will be come a staple in your optimization strategy.
NB: This is an analysis by Chloë Lind, managing consultant at Maxymiser. You can follow her on Twitter here. It appears here as part of Tnooz’s sponsored content initiative.