December 2024 Summaries

In order to effectively implement AI within an organization, it's crucial to consider governance capabilities, tech stack, and data quality to tailor a plan according to readiness levels. This involves building strong processes in place for managing AI use, ensuring compliance with industry standards, and fostering cross-team collaboration. A connected tech stack is also essential, emphasizing interoperability, adaptability, and scalability. Moreover, high-quality data that's clean, organized, and accessible is vital, as poor data can lead to poor decision-making and ultimately worse business outcomes. By focusing on the three pillars of AI readiness - process, technology, and data - organizations can lay a strong foundation for AI success and reap its transformative benefits.

AI agents are systems that can independently make decisions and take actions to accomplish specific tasks, operating using four basic steps: perceiving, analyzing, decisioning, and action. These agents have historically been difficult to implement due to complex machine learning models or rules-based programming, but advancements in data availability, computational power, and foundational models have made them more accessible. The key to effective AI strategy lies in a strong data foundation, with organizations needing to address three dimensions of data readiness: data quality, connectivity, and collaboration. By leveraging AI agents, organizations can drive value through increased productivity, enhanced decision-making, and improved customer experiences, making it a competitive advantage in today's fast-paced technological landscape.

Salesforce is a leader in CRM, and by connecting it with Census and utilizing GPT Columns, businesses can supercharge their workflows with AI-powered insights. This enables use cases such as automating fit scoring in RevOps, building an ideal customer profile analysis and account scoring system, creating product usage summaries to empower GTM teams, and leveraging AI to deliver personalized recommendations, prevent churn, and craft hyper-personalized marketing campaigns.

I learned the importance of selecting more efficient language models for tasks that don't require heavy processing power, and being selective about testing and processing to avoid unnecessary re-processes and costly mistakes. I also implemented safety measures such as budget alerts and restricted access to expensive models to prevent financial disasters. Additionally, I adopted a simple monitoring system using Google Sheets to track issues and identify areas for improvement.

Braze's pricing model can be challenging due to its unpredictable structure based on the volume of Data Points consumed. This has led to scalability issues for many companies. Census is a tool that helps users accurately predict Braze API costs and optimize their Data Point spend, especially for those with large user bases or complex personalization needs. It offers granular diffing and only syncs specific Data Points that have actually changed, minimizing data point consumption and reducing costs associated with Braze's pricing model. Census also provides features such as data point optimization, batching API requests, and sync dry runs to help users estimate the number of Data Point writes for every sync, making it easier to use Braze despite its complex pricing structure. By using Census, businesses can supercharge their Braze experience, minimize costs, and monitor data flows without worrying about unexpected issues.

The text discusses the potential of using AI, specifically Census GPT Columns, to identify customers who have the potential to spend more before they even realize it themselves. It provides a step-by-step guide on how to use this technology in a fitness business context, where the goal is to identify which lower-paying customers are exhibiting premium customer patterns and could potentially upgrade their membership plans. The process involves understanding premium customer characteristics, creating an Upsell Score based on these traits, and operationalizing the score by sending targeted marketing communications to individuals with high scores. This approach can be applied across various industries and business models to transform data into actionable insights and improve customer targeting for upselling opportunities.

The text discusses using artificial intelligence (AI) to identify high-potential customers, particularly those who are likely to become premium customers. This is achieved by analyzing customer traits and behaviors against known indicators of premium preferences. A scoring system, called an "Upsell Score," is created to automatically identify customers showing premium patterns. The AI Columns tool is used to analyze these traits and create the Upsell Score, which can then be operationalized in marketing campaigns using 200+ marketing automation and advertising tools. This process enables businesses to precisely target customers whose traits signal readiness for premium offerings, leading to more meaningful conversations and better understanding of customer segments.

To prevent customer churn and maintain high net dollar retention, it's crucial to keep customers engaged. Identifying at-risk customers and intervening at the right moment can be challenging. This guide explains how to use Census's GPT Columns to create data points like an engagement score and trend, which can help identify valuable customers who are at risk of disengagement. These insights can then be used to power re-engagement campaigns in platforms such as Braze. The process involves identifying high-value customers, understanding their engagement levels, defining a dataset with relevant events, setting up GPT Columns for analysis, and syncing the data points to Braze or other marketing platforms.