Case Study: G5 achieves 95%+ lead-scoring accuracy and $1.5M/month savings with H2O.ai Driverless AI

Increasing Effectiveness of Real Estate Marketing with H2O.ai

G5, a marketing optimization company serving more than 7,000 real estate properties, faced very low productivity from call-based leads (only 14% were productive) and lacked dedicated data science resources to improve conversion rates. To address this, G5 partnered with H2O.ai and used H2O Driverless AI (along with H2O Word2Vec and H2O-3) to build machine learning models from 100,000 call transcripts and integrate predictions into its Intelligent Marketing Cloud.

H2O.ai’s solution automated feature engineering, model training and scoring (deployed via AWS Lambda), enabling G5 to identify high-quality leads and present leasing agents with prioritized lists. As a result, lead-scoring accuracy rose to over 95%, qualified lead connections climbed to 85%, model development time fell by 80%, and customers realized about $1.5M/month in savings—improving agent effectiveness, satisfaction, and retention.

G5

Martin Stein

Chief Analytics Officer