Surviving the Reset: A Framework for AI Competitive Advantage (welovesota.com)

A new valuation framework addresses why traditional SaaS metrics (CAC/LTV, retention, steady growth) fail for AI companies and proposes a four-level hierarchy—Technical Foundation, Product Depth, Market Position, and Sustainability—to evaluate what creates durable value. It emphasizes three technical pillars: Domain Delta (real-world performance on buyer-relevant metrics under production constraints), Velocity Dynamics (rate of meaningful improvement versus the frontier), and Effective Data Advantage (exclusive, high-quality data pipelines or synthetic-data strategies). Product depth is judged by job coverage, how deeply a product embeds in workflows (peripheral → foundational), and execution autonomy (from suggestions to full automation). Market positioning reads capability thresholds, vertical vs. horizontal play, and expected market concentration. Sustainability is tested by scenarios like a GPT-N reset, net switching friction, and new-user win rates. The framework’s practical implications: single moats erode quickly—sustained advantage requires layered, compounding defenses (fast release cadence, defensible data, and workflow integration). It gives archetype-specific benchmarks (foundation models must match 3–4 month frontier cycles; verticals can tolerate 6–12 months; apps need weekly/monthly UX velocity; infra can be slower but must lock developers). Real-world examples (Jasper’s rapid devaluation post-ChatGPT, Raidium’s Curia radiology model, PriorLabs’ synthetic TabPFN, Perplexity’s daily iteration) show how displacement risk and capability-driven market creation demand new KPIs and sanity checks for investors and builders.