$200M/Year Portfolio Owner: Here's How We're Using AI Across our Businesses

Quick Overview

The portfolio owner, who manages a $200M/year portfolio, leverages AI by training custom GPTs as specialized agents or roles within their acquired businesses, using proprietary intellectual property to unbottleneck processes and transform operations, while simultaneously noting that businesses selling anything other than a direct end result, like information products, face obsolescence.

Key Points: The speaker experienced a significant slowdown in deal-making in 2024 due to marketplace uncertainty, though they anticipate a potentially biggest deal this year that could increase the portfolio size by at least 50%. Their acquisition model targets businesses that have been flat or down, as they are more open to new ideas and bringing in partners for sweat equity, contrasting with traditional private equity seeking growth. The core process for turning around a business starts with visually mapping how customers happen or how fulfillment occurs to identify demand or supply constraints, operating under the Theory of Constraints to debottle bottlenecks. AI massively streamlines delivery by training custom GPTs on proprietary video content, allowing an advisor to interface with the client while the AI asks questions, achieving 80-90% accuracy in extracting necessary information. The speaker strongly believes information businesses are facing a crisis, stating that if they are not pivoting, they should decide on a "slow and profitable death," as customers now prioritize results over training courses, exemplified by Digital Marketer's course sales dropping to 20% of prior levels. They are building AI agents trained on specific expertise (like an 'email marketing expert') using proprietary IP, rather than just task-based GPTs, to effectively create an AI 'role' or 'employee' to supplement existing teams. In portfolio companies, the first step implemented after acquisition is an AI audit against the value engine process maps, heavily encouraging at least one primary initiative per quarter to be dedicated to AI optimization at bottleneck stages.