80 hours of maintenance time saved per week with Vertex AI managed services
6 months of model development time saved with automated feature engineering
Up to 50% increase in ROAS with causal AI models built with Google Cloud
Decision-making platform Churney needed a way to develop and deploy its sophisticated causal AI models quickly, efficiently, and at scale. With the managed services and automations of Vertex AI, Churney can test hundreds of models simultaneously, helping businesses increase their return on advertising spend.
Decision-making platform Churney needed a way to develop and deploy its sophisticated causal AI models quickly, efficiently, and at scale. With the managed services and automations of Vertex AI, Churney can test hundreds of models simultaneously, helping businesses increase their return on advertising spend.
Building a scalable decision-making platform with Google Cloud
As a number of internet browsers phase out the use of third-party cookies due to privacy concerns, businesses must learn to use first-party data effectively to drive growth. AI-driven decision-making platform Churney addresses this challenge by helping businesses understand customers’ interactions with their websites and apps to make smarter decisions. It does this using causal AI, a powerful new approach that goes beyond finding patterns in data to understand the cause-and-effect behind those patterns, giving marketers deeper insights to help them optimize their marketing strategy and spend.
With causal AI, Churney enables businesses to use first-party data to understand why customers take certain actions and make accurate predictions about what they’ll do next. In particular, it predicts the lifetime value of customers, enabling businesses to focus resources on high-value users for a better return on advertising spend (ROAS). Churney’s causal AI models also allow businesses to replace traditional one-size-fits-all A/B testing with a personalized approach. By predicting what messaging individual users will respond to best, Churney enables businesses to target users with relevant messages, improving customer retention.
Causal AI models depend on a vast amount of data. Being able to process and work with that data in a secure and cost-effective way were key considerations for Churney as it searched for a cloud platform to build its decision-making platform. “The need for secure, cost-efficient data handling made Google Cloud an easy choice,” explains Noy Rotbart, Co-founder and CEO of Churney. “BigQuery can run scalable data warehouse operations out of the box. With its efficient pricing model, you can build your business around it.”
The need for secure, cost-efficient data handling made Google Cloud an easy choice. BigQuery can run scalable data warehouse operations out of the box. With its efficient pricing model, you can build your business around it.
Noy Rotbart
Co-founder and CEO, Churney
Saving months of AI maintenance and monitoring work with Vertex AI
With BigQuery as its unified data platform, Churney uses Vertex AI to access data and train and deploy the causal AI models powering its decision engine. Its serverless infrastructure, managed services, and automations streamline Churney’s MLOps and model development workflows, saving considerable maintenance and model development time.
“Without Vertex AI we’d need to spend 80 hours a week maintaining our services,” Rotbart says. “In just three clicks, Vertex AI gives us Model Monitoring off the shelf. That’s irreplaceable. We also would have needed a two-person team to monitor data drift. Instead it’s one button with Vertex AI. Every feature Vertex AI gives us would have taken up to three months to develop ourselves.”
These cost and time savings allow Churney to channel resources into its data science team to develop and refine more accurate models for better predictions and more effective personalization. The company can also pass on efficiency savings to customers in the form of more affordable services.
In just three clicks, Vertex AI gives us Model Monitoring off the shelf. That’s irreplaceable. We also would have needed a two-person team to monitor data drift. Instead it’s one button with Vertex AI.
Noy Rotbart
Co-founder and CEO, Churney
Using Churney’s causal AI models to predict customer lifetime value, businesses can focus resources on their highest-value customers. “Our customers are consistently seeing a 20% increase in ROAS with our causal AI predictions,” says Rotbart. “One customer saw 2.4 times the return on investment they previously saw. With the help of Vertex AI, we’ve made a money machine for our customers.”
Perfecting personalization with the ability to test hundreds of models simultaneously
Churney also uses Vertex AI to train and deploy causal AI models to help businesses retain customers over the long term. Traditionally, businesses use A/B testing to see which message converts best. With its causal AI models, however, Churney enables businesses to simulate how users would respond to different messages. This means they can identify the best treatment for each customer, such as a tailored discount or a personalized message, to increase engagement and maximize retention.
Vertex AI gives us the ability to speed up experimentation and test hundreds of models simultaneously, enabling significant model improvement.
Brian Brost
Co-founder and CTO, Churney
Vertex AI's robust infrastructure and automation capabilities help Churney streamline its causal AI model development and training. “Vertex AI gives us the ability to speed up experimentation and test hundreds of models simultaneously, enabling significant model improvement,” says Brian Brost, Co-founder and CTO of Churney.
Churney uses the Vertex AI Workbench, Pipelines, and Feature Store to automate feature engineering, extracting meaningful signals from semi-structured data such as logs of user activity, gathered with customers’ consent.
This automation allows Churney to uncover complex patterns in user behavior and build increasingly sophisticated causal AI models for accurate outcome predictions, paving the way for customers to be shown highly relevant personalized treatments.
“Vertex AI has allowed us to focus on automating our feature engineering and build thousands of micro features to refine our causal AI models,” Rotbart explains. “Without that, it would have been difficult to have scaled.”
By personalizing treatments with Churney’s causal AI, businesses can optimize campaigns to improve results and build lasting customer relationships. A recent Churney customer saw a 31% increase in ROAS and a 36% lift in repeat purchases, while another saw a 50% increase in day 30 ROAS, highlighting Churney’s effectiveness in optimizing long-term campaign performance and maximizing the value of acquired users.
Churney is now building on its success in predicting single customer interactions by helping businesses personalize entire customer experiences for even greater ROAS. The company is also deepening its collaboration with Google Cloud to bring its causal AI-driven decision-making platform to joint customers. “Our collaboration with Google Cloud will be a significant step forward for Churney, allowing us to deliver even greater value to more customers,” says Rotbart.