AI SaaS Strategy: TAM Expansion and Revenue Stacking
Atlassian co-founder Mike Cannon-Brookes and VCs analyze the AI-driven shift in software markets. Key themes include the obsolescence of static TAM models, the critical distinction between input-constrained and creation-driven domains, and the strategic imperative for SaaS leaders to demonstrate AI-driven value acceleration.
The Shift from Static TAM to Dynamic Revenue Validation
The prevailing narrative that software is dying is a misinterpretation of market dynamics. As Atlassian co-founder Mike Cannon-Brookes argues, the industry is not shrinking but restructuring. The core strategic error lies in applying static Total Addressable Market (TAM) models to an era of exponential capability expansion. When AI providers like Anthropic and OpenAI project revenues that exceed the entire legacy software market, the assumption must be TAM expansion, not zero-sum displacement. Investors must abandon the habit of capping growth based on historical budgets and instead let revenue velocity validate the new market ceiling.
Revenue Stacking and the Illusion of Market Share
A critical analytical blind spot is the failure to account for revenue stacking. AI revenue figures often include costs passed through to cloud infrastructure and model providers. For instance, a SaaS vendor’s spend on AI models contributes to the revenue of both the model provider and the underlying cloud infrastructure. This layering means that headline AI revenue numbers do not represent a direct substitution for application-layer software budgets. Understanding this stack is essential for accurate valuation and competitive analysis, as it reveals that the total addressable spend is significantly larger than the sum of individual vendor revenues.
Input-Constrained vs. Creation-Driven Domains
The impact of AI varies drastically by domain. In input-constrained areas such as legal services and customer support, AI primarily drives efficiency and headcount reduction, leading to potential seat shrinkage. Conversely, in creation-driven domains like software engineering, AI accelerates output, leading to increased demand for collaboration and tracking tools. This distinction explains why product and engineering tools are seeing growth while traditional support software faces pressure. Companies must identify whether their core value proposition lies in reducing input costs or enabling greater output creation.
Strategic Imperatives for SaaS Leaders
Public SaaS companies face a unique challenge: balancing short-term financial metrics with long-term AI investment. Unlike private peers who can burn capital for growth, public firms must demonstrate immediate ROI while funding R&D. The path forward requires a clear narrative on how AI integration drives customer value, not just cost savings. Leaders must focus on agentic automation that executes business processes, moving beyond simple chatbots to deliver tangible operational improvements. The winners will be those who can prove that AI adoption leads to accelerated revenue growth and expanded market share, rather than merely defending existing budgets.
Key insights
-
Static TAM models are no longer valid for AI-driven software. The market is expanding due to increased productivity and new use cases, not just reallocating existing budgets.
Impact: Investors and operators who rely on historical TAM caps will undervalue high-growth AI companies and miss significant market opportunities.
-
AI revenue figures are often inflated by revenue stacking, where costs flow through multiple layers including cloud infrastructure and model providers.
Impact: Misinterpreting stacked revenue as direct market share can lead to flawed competitive assessments and incorrect valuation multiples.
-
Domains are either input-constrained (e.g., legal, support) or creation-driven (e.g., engineering). AI reduces headcount in the former but expands tooling demand in the latter.
Impact: Companies in creation-driven domains are positioned for growth, while those in input-constrained domains must pivot to efficiency-focused value propositions.
-
Public SaaS companies face a capital allocation tension between meeting short-term EPS targets and investing in long-term AI R&D.
Impact: Leaders who fail to communicate a clear AI strategy to investors risk undervaluation, while those who balance both can sustain long-term growth.
-
The value of AI in enterprise software is shifting from answering questions to executing actions through agentic workflows.
Impact: Products that automate end-to-end business processes will command higher premiums and drive stronger customer retention than simple chat-based interfaces.
Action items
-
Re-evaluate market sizing models to incorporate dynamic revenue growth rather than static TAM caps. Focus on revenue velocity as the primary indicator of market potential.
Impact: This approach will align valuation with actual market expansion, reducing the risk of undervaluing high-growth AI companies.
-
Map your product’s domain as either input-constrained or creation-driven. Adjust your value proposition and go-to-market strategy accordingly.
Impact: Tailoring strategy to the domain type will help you target the right customer pain points, whether it’s cost reduction or output expansion.
-
Develop a clear narrative for investors on how AI investment drives long-term value. Quantify the ROI of AI R&D in terms of customer acquisition and retention.
Impact: A compelling AI narrative will help public companies maintain investor confidence and secure the capital needed for sustained innovation.
-
Prioritize the development of agentic workflows that execute business actions, not just provide information. Focus on automating end-to-end processes.
Impact: Agentic automation will differentiate your product in a crowded market and drive higher customer lifetime value through tangible operational improvements.
-
Analyze your revenue stack to understand how much of your AI spend is passed through to infrastructure and model providers. Use this data to refine your cost structure and pricing.
Impact: Understanding the revenue stack will help you optimize margins and position your product more effectively in the AI value chain.
Quotes
“I think we have to give up on TAM. I think we just have to let the revenue show us the path to TAM.”
“I think every category that I know of outside of engineering and product is at existential risk of shrinking seats.”
“The idea that software as a category is dead is ludicrous to me.”