As a data-driven agency, Merkle | Periscopix’s optimisation process always starts with a deep analysis of data to provide insights into onsite customer behaviour. We began by analysing data in Google Analytics, as well as the client’s CRM data, uncovering some key insights:
- Users had a tendency to browse and purchase within a certain brand, without jumping from brand to brand, indicating a strong brand affinity;
- 40%+ of these users tend to land on the homepage when they return to the site;
- Less than 2% of these users interacted with the homepage carousel.
Our next step was to understand the “why” behind the data. We performed a customer journey analysis to identify the barriers customers face when using the site; almost half of these user journeys start on the homepage. While scrutinising this landing page, we were able to expose a few more pieces of insight:
- Users landing on the homepage are presented with nine different brands/promotions above the fold;
- With 30+ brands onsite, the likelihood that the brand the user has an affinity to is shown above the fold is fairly low – potentially explaining why carousel interactions are lower than the industry average.
So what is the problem here? We know the user, we know their intent, we understand their behaviour from the data, but the site is still serving the same experience to all users.
Personalised Experiences to All Users
The opportunity here became very clear. And our hypothesis was born:
If we serve a more personalised experience onsite by promoting the same brand that each user had previously interacted with on the homepage carousel, then we should see an increase in the number of transactions related to these product categories because we know that users have a strong brand affinity and priming them with their specific brand will help them move on in their journey.
To create a more personalised homepage, we replaced the homepage carousel images with brand-specific images of the three top-performing brands on the site (Castelli, Specialized, Assos). With the GA Audience targeting feature in Optimize 360, we were able to serve these bespoke experiences to the appropriate subsets of users. We created three distinct audiences in GA of users who had previously interacted with, or purchased from, these three brands and used them as targeting rules in Optimize.
For example, if a user had previously interacted with or purchased from the Specialized brand, when they returned to the homepage, they would see only Specialized items in the carousel.
Personalising the homepage proved to be very successful. All three treatments outperformed the original, with a 90%+ probability to beat the baseline.
From click through rates on the carousel, to product detail views, add to cart and transactions of the respective brands, we saw a significant uplift. Overall, we saw a 32% increase in eCommerce conversion rates across the three experiments, as well as a 28% increase in revenue.
The image below shows the average improvement of variants compared to the original across the three experiments.
Improvement of Personalised Variation Compared to Original
At Merkle | Periscopix, we treat Conversion Rate Optimisation as an iterative process, that never ends. We are pushing boundaries to continuously understand the end user and present them with more personalised experiences onsite to help them achieve their goals.
For Sigma Sports, we are looking to further enhance their user experience by using GA as data source to drive optimisation and personalisation with Google Optimize 360.