AI-Driven Email Optimization: How InXiteOut Increased B2B Offer Conversions by 50% 

Client Context 

Our client, a Fortune 500 CPG company, leveraged B2B email campaigns to send targeted sales offers to its network of retailers. These campaigns were executed on Mailchimp, relying on the platform’s automated suggestion for the “best” send day and time. 

Despite executing multiple campaigns, the sales uplift remained unsatisfactory as retailers were not engaging with the emails. The key metrics across campaigns were: 

• Open Rate: ~14% 
• Click Rate: ~8% 
• Offer Conversion Rate: ~3% 

Low engagement led to low ROI for the email campaign, thereby reducing overall confidence in the channel as a sales lever. 

The client partnered with InXiteOut to enhance campaign engagement and B2B offer conversions by developing a data-driven strategy to optimize send date and time. 

The InXiteOut Approach 

Our hypothesis was that Mailchimp’s recommendations were failing because they were one-dimensional. The platform's algorithm could track email data (such as when an email was opened), but was blind to the most critical business data: the retailer's actual purchase history and e-commerce browsing behavior. 

Our solution was to build a custom AI engine that could see this entire data ecosystem. This involved three key stages: 

Creating a 360-degree retailer view 

Before we could build a model, we had to break down the client's data silos. We created a centralized, harmonized data lake that combined disparate data sources into a single, unified view for each retailer. This included: 

• E-commerce Data: Past purchase history, order frequency, average order value, and offer redemption patterns. 
• Web Analytics: Google Analytics/GTM data on portal browsing patterns and content engagement. 
• Campaign Data: Historical email engagement (opens, clicks) from Mailchimp. 

This unified dataset was the essential foundation for an accurate predictive model. 

Building a dual-objective AI engine 

With this 360-degree view, we enriched the data with curated features (e.g., campaign type, brand involved, past behavior) to build a far more sophisticated AI model. 

The key difference was that our model was built with two objectives, not just one: 

• Optimize Engagement: Predict the send date and time most likely to maximize open and click rates. 
• Optimize Conversions: Go beyond simple engagement to find the send time patterns that actually led to an offer conversion. 

This dual-objective model allowed us to find the "sweet spot" that balanced getting a retailer's attention (opens) with the ultimate business goal: driving a sale.  

Validating success with smart A/B testing 

Finally, we validated our model's recommendations by running a head-to-head A/B test in a pilot market. One group of retailers received emails based on Mailchimp’s recommendations, while the other group received them based on our new AI model. This allowed us to directly measure the performance lift. 

End-to-end productionization 

Following the successful pilot, we productionized the entire solution. We built an automated pipeline that continuously ingests new data, re-trains the model, and feeds the optimal send-time recommendations for every individual retailer directly into the Mailchimp platform. 

This transformed a one-time analysis into a scalable, self-learning system that permanently enhanced the client's marketing operations. 

Technology stack used 

• Azure ETL and ML Ecosystem (Data Lake, Data Factory, Databricks)     
• MailChimp 
• Power BI  

Benefits Delivered  

The solution delivered significant and consistent results across all key metrics: 

• +50% Increase in Offer Conversion Rate: Directly met the primary business goal by optimizing the campaigns for actual sales, not just email engagement. 
• +30% Boost in Email Open and Click Rates: Decisively outperformed the platform's generic recommendations, capturing retailer attention at the right time. 

By productionizing this solution, InXiteOut transformed the client's email channel from a low-performing asset into a predictable, high-ROI sales driver.