Andreessen Horowitz partner David George has delivered a blunt message to the software industry: the era of comfortable mediocrity is over. In a detailed analysis, George argues that public markets have already repriced the software sector, and companies now face only two viable paths to creating lasting shareholder value. The stakes are high, and the timeline is tight. What Are the Two Paths Forward for Software Companies? George outlined two distinct strategies that software companies must pursue over the next 12 to 18 months. The first path requires accelerating revenue growth by 10 or more percentage points year over year through genuinely new AI-native products. The second path demands rebuilding the company to achieve 40 percent or higher true operating margins, ideally reaching 50 percent, while treating stock-based compensation as a real expense rather than a carve-out. The critical insight is that these paths are not mutually exclusive in theory, but in practice, George suggests that companies need to commit to one or the other within the 12-to-18-month window. Anything caught between these two poles will face what he calls "no-man's land": persistent growth pressure, ongoing dilution, and multiple compression that destroys shareholder value. Public software companies have already experienced the first half of this transition. Growth has rolled over and valuations have compressed significantly. However, true profitability has largely failed to materialize in most cases. While free cash flow and GAAP margins have improved somewhat, the sector remains stuck in a difficult middle ground when stock compensation is treated as a genuine expense rather than excluded from calculations. How Should Software Leaders Rebuild Their Organizations for AI? George provides a detailed roadmap for companies pursuing the growth path through AI-native products. The strategy begins with identifying and empowering a small core team of exceptional leaders who will drive the transformation, regardless of their current organizational rank. George emphasizes that somewhere within every organization exist approximately five people capable of delivering 100 times the value previously thought possible. The first phase involves these five leaders conducting intensive information-gathering projects across the organization. This includes capturing processes around high-value workflows, harvesting standard operating procedures, support tickets, transcripts, requirements documents, policies, customer relationship management notes, support logs, event data, and approval paths. The goal is to create a living context layer rather than static documentation, treating documentation as product infrastructure with clear evaluations around accuracy, exception handling, latency, and cost. Steps for executing this organizational transformation include: - Leadership Assessment: Watch vice presidents closely during the information-gathering sprint to determine who is committed to the transformation and who should be replaced with AI-native leaders and high-performing team members. - Executive Restructuring: After one month, have difficult conversations with executives who are not aligned with the new direction, then replace them with the team that completed the information-gathering sprint and other AI-native up-and-comers who have earned their opportunity. - R&D Reorganization: Allocate 50 percent of research and development resources to net-new AI products using four-person pods that collapse design, product, and engineering into single working units, starting code development on day one. - Product Manager Focus: Ensure all best product managers remain customer-facing and focused on pure product discovery, unblocked by legacy systems and processes. - Engineering Architecture: Keep best engineers in the central engineering organization reporting directly to the chief technology officer, prioritizing the company's core architecture evolution while ruthlessly prioritizing new product development. - Decision-Making Speed: Get the reformed executive team to dedicate at least one full day per week exclusively to unblocking designers, product managers, and engineers, making hard choices every single week to maintain momentum. George emphasizes that this is not the "weak form" of restructuring, which merely trims the organizational edges while leaving the core machine intact. Instead, he calls for the "strong form," which is a complete redesign of how the company operates. The 8 to 10 percent layoff headlines that have dominated recent news cycles no longer count as meaningful transformation. Why Is the Pricing Model Shift So Critical to Success? A crucial element of this transformation involves fundamentally rethinking how software companies monetize their products. Traditional seat-based pricing models, where companies charge per user, are becoming obsolete in an AI-driven world. Instead, George argues that companies must transition to token-based or per-use pricing models where agents and autonomous systems can consume and pay for products independently. The business reality driving this shift is straightforward: customers' first and most obvious source of AI savings is labor efficiency, which means they will look to reduce the number of seats they purchase. The new growth, by contrast, will increasingly come from tokens, consumption-based metrics, automations, outcomes, and machine-driven workflows. If a company is not positioned in the token path, it is not standing in the fastest-growing part of the customer's budget. George notes that companies do have time to make this transition, as seat-based pricing will not disappear overnight. However, the challenge cannot be hand-waved away with vague promises. If an AI agent cannot autonomously consume and pay for a product, the company probably has not achieved true product-market fit in the AI era. "The budget for new spend is there. You can do this," stated David George, emphasizing that the financial resources exist for companies willing to make bold moves. David George, Partner at Andreessen Horowitz What Happens to Companies That Cannot Execute This Transformation? George acknowledges that not all companies are positioned to pursue the high-growth path through AI-native products. Some organizations may assess their options and conclude that no credible line of sight exists to winning through accelerated growth. For these companies, the alternative path of achieving 40 to 50 percent operating margins becomes the only viable strategy. However, for companies that do have a realistic path to growth and successfully execute the 12-month sprint, the payoff is significant. They will emerge as focused, accelerating organizations with refreshed leadership teams and what George calls a "refounding moment." This moment can energize teams and provide momentum for years to come. The broader message from Andreessen Horowitz is clear: the software industry's comfortable middle is disappearing. Public markets have already sent this signal through repricing. Companies that fail to commit decisively to either the growth path or the profitability path will find themselves trapped in a value-destroying middle ground with no clear way out. The time for bold management decisions is now, and the window for action is measured in months, not years.
