Asics came to Keydabra wanting to learn what behavior and what specific content drives higher engagement so they could optimize their site and convert more traffic into sales.
Using only data captured by Google Analytics on November 24, 2017 (“Black Friday”) and January 30-31, 2018, Keydabra’s proprietary algorithm generated a probability-of-conversion (make a purchase) for each customer, as well as an engagement score for all website visitors. Four Keydabra modules utilizing 25 different machine learning algorithms were used to identify the hidden behavior patterns of website visitors.
Finding the Ideal Customer
Keydabra identified the demographic factors that had the highest potential to convert to customers. For Asics’ specific goals, that meant females in New York, aged 18-24, who utilized Google to arrive at the footwear and sports equipment company’s website. Keydabra also isolated three behaviors that indicated higher levels of customer engagement: using the page search function, using the store locator function, and viewing specific site content.
These factors allowed Asics to optimize their site and target their ideal customers to increase their engagement score and enable a higher conversion rate and increased ROMI.
Recommendations for Continued Improvement
Beyond identifying who to target and what content and behavior drive higher levels of engagement, Keydabra interpreted the data to devise solutions for improving the overall experience of website visitors. Recommendations included improving the search function and algorithm, identifying keywords that were vital for higher engagement rates, remarketing based on customers’ engagement scores, identifying products that should receive enhanced marketing and promotions, and offering discounts based on propensity to buy.
Keydabra also isolated data showing that VSSL ran the risk of losing 52% of current customers who were in “hibernating mode.” It was recommended that these customers be reactivated through targeted email campaigns.