These examples highlight the wide range of possibilities AI can unlock across your day-to-day work and will help spark new ideas for how you can use AI in your own marketing.
ThirdLove uses AI to orchestrate smarter omnichannel experiences
ThirdLove is a women’s intimates brand known for creating bras that fit each customer’s unique preferences. It’s no surprise that they want their customer experience to feel just as personalized. With a customer base spanning multiple demographics, product needs, and shopping behaviors, the team needed a way to tailor communication across email and text messaging without overwhelming subscribers or relying on guesswork.
Goal
ThirdLove wanted to ensure each customer received messages on the channel they were most likely to engage with (whether email, text, or both) while reducing unnecessary sends and improving overall engagement.
Strategy
ThirdLove decided to use Klaviyo’s channel affinity, which relies on machine learning to predict how likely each customer is to engage with their next message on a specific channel. By analyzing real engagement signals like opens, clicks, and recent subscription activity, the model identifies each customer’s preferred channel at scale. This allowed ThirdLove to build cross-channel flows and campaigns that dynamically deliver messages where customers are most likely to engage.
Results
In Q2 2025, ThirdLove saw substantial quarter-over-quarter growth in flow revenue, along with a notable increase in revenue per campaign. SMS, in particular, delivered a 15x return on investment, demonstrating the impact of AI-powered personalization at scale.
Criquet Shirts uses AI to shorten their experimentation cycle
Criquet Shirts is a menswear brand known for its vintage-inspired polos and loyal customer base. When the team noticed lower engagement on email, they wanted to run targeted experiments to find the right segments to message. However, with large datasets spread across tools, building segments manually took weeks, slowing their ability to test and optimize. They needed a faster, more flexible way to analyze their data so they could run experiments at the speed their business required.
Goal
Criquet Shirts wanted to identify high-performing email segments more quickly and improve engagement through faster experimentation.
Strategy
Criquet Shirts used Orita AI to analyze engagement data and generate test segments automatically. Instead of querying large datasets with SQL or manually stitching reports together, the team could ask questions in plain language and receive ready-to-use segments within minutes. This made it possible to test multiple segmentation strategies in a much shorter time frame.
Results
Manual data analysis and segment creation previously took weeks. With AI, the team could generate segments and launch tests the same day, shortening each experiment from 3 weeks to 1 week. Faster experimentation led to a 16.7% increase in click rates, which helped protect the brand's deliverability. It also helped reduce their email segment size from 250,000 to 80,000 recipients while keeping revenue constant.
Underground Ecom uses AI to turn data into actionable insights
Underground Ecom is a Klaviyo Elite Master partner known for helping brands build highly profitable customer relationships through data-driven retention marketing strategies. When one of their clients, a US-based perfume manufacturer, experienced a drop in email engagement, Stefan Milicevic and his team needed to understand why quickly. But with limited bandwidth and the added complexity of managing fast-changing inventory data, manual analysis wasn’t a realistic option.
Goal
Underground Ecom wanted to improve email click rates and revenue for their client by gaining a deeper, data-driven understanding of how the brand’s audience engaged with different types of campaigns.
Strategy
Stefan’s team used Underground Insights, their proprietary AI-powered reporting tool, to analyze historical email performance at scale. Instead of manually reviewing individual sends, the tool surfaced engagement patterns across campaigns. The analysis revealed that engagement was consistently stronger in promotional emails than in brand-building ones. Based on this insight, the team updated the brand-building email template to incorporate proven promotional elements, such as removing the header navigation and moving the primary call to action above the fold.
Results
By using AI to analyze large volumes of data and surface meaningful patterns, Underground Ecom quickly identified the root cause of the engagement drop while reducing time spent on manual analysis. This freed the team to focus on strategy and optimization. After updating the email template, the client saw an 8% increase in click rates, with performance continuing to improve over time.