How Lovisa Uses AI to Deliver Hyper-Personalized eCommerce Experiences

In today’s age of omnichannel, consumers expect a personalized experience whether they’re shopping online or in-store. 

In fact, research from Merkle shows that 86% of respondents are likely to enjoy personalized offers based on their interests and browsing or purchase history, while a whopping 90% express a willingness to share more data about themselves if they have a positive experience with a brand. These two factors feed into each other, allowing merchants to leverage their zero- and first-party customer data to deliver tailored experiences. 

Since launching its first brick-and-mortar store in 2010, the trendy and affordable jewelry brand Lovisa has expanded to more than 30 countries and made the foray into eCommerce to reach even more consumers. To scale their online experience to match the growth of retail, the brand has doubled down on an AI-powered hyper-personalization strategy to build better customer journeys and drive higher AOV.

Matthew Stuckings, CRM and Data Manager for Lovisa, recently sat down with Marjorie Li, AfterShip’s Global Marketing Director at Online Retailer — Australia’s largest eCommerce conference and expo in Sydney, to discuss the brand’s strategy, what they uncovered in their data, and what’s next for their business. Here are the highlights. 

The importance of knowing your customer

“In the eCommerce game, there are so many unanswered questions — so many options that you have with technology — what's the strategy you're going to use? You need to find a starting point,” said Stuckings. “So for us, we used the preexisting data we had from our CDP to give us those insights to think of what the strategy to start with would be.”

By digging into the data, the brand uncovered a trove of interesting insights, including:

• Regardless of the product type in a customer's first order, their second order almost always included earrings. This is likely because earrings cannot be returned, and the customers need to establish trust in the brand before choosing to buy them.
• No matter what product type a customer purchased in their first order, their second order was very likely to contain exactly the same product type.

By leveraging these insights along with AI-powered algorithms, Lovisa was able to build out a personalization strategy to drive higher AOV and provide a tailored experience to every customer. 

Human intelligence + artificial intelligence = a winning formula

Lovisa uses AfterShip Personalization to power product recommendations throughout the customer journey. By using a mix of AI algorithms and merchandising/custom rules, the brand can display conversion-driving widgets on its product pages and at checkout. 

It’s not enough to just set it and forget it, however. Lovisa performs extensive A/B testing to determine the highest-performing widgets. For example, the brand tested “similar products” against “frequently bought together” widgets, and found that similar products won by 30-40%.

“I think that the mixture of personalization algorithms that AfterShip offers and putting your own merchandising rules into that — that's where the AI meets the road with interesting things that are actually going to deliver results,” said Stuckings. “And looking at the data, even a small AOV boost can have a big impact at scale.” 

Here are some of the results the brand has seen with AfterShip Personalization:

• 5% lift in AOV with personalization vs. without
• Personalization-attributed revenue accounts for 3% of Lovisa’s total revenue
• 21x ROI with AfterShip Personalization

The vast majority of the brand’s personalization revenue — 74% — comes from product page widgets. However, personalized recommendations on Lovisa’s checkout also contribute to revenue, and that’s without any offers or discounts to incentivize conversion. 

The future of Lovisa’s hyper-personalization

Lovisa plans to continue A/B testing personalization widgets across its online store to further boost AOV and revenue. The brand is also beta testing AfterShip’s new Discovery AI, which uses 500 million transaction data points and an affinity database to enable personalized search results and recommendations to help shoppers find the most relevant products for them.

“What this looks like in practicality for us is, people that are browsing and looking at, say for example, blue feathered earrings, we know that they're more likely to purchase from the Bohemian collection,” said Stuckings. “Even if you don't have a lot of data about the customer you've got a lot of data when you have the processing power of the large language models to create these kinds of affinity databases. We can look out for these touchpoints and behavioral patterns and then recommend products based on that.”

Lovisa’s success shows that with the right technology and the right team, eCommerce merchants can deliver game-changing personalized experiences that drive conversion, higher AOV, and repeat business. 

Are you ready to shift your personalization into hyperdrive? Learn more about AfterShip Personalization or request a demo now.