AI products with a Product-Led Growth motion face a fundamental flaw in their unit economics. Customers expect predictable SaaS-like pricing (e.g., $20/month), but the company's costs are usage-based. This creates an inverse relationship where higher user engagement leads directly to lower or negative margins.

Intense demand for AI tokens is outstripping compute supply, making flat-rate SaaS pricing unsustainable. Companies like GitHub are now shifting to usage-based billing to cover escalating inference costs, marking a fundamental change in how AI products are sold and signaling a broader industry trend.

Pure value-based pricing (e.g., per seat) fails for AI products due to unpredictable token costs from power users. Vercel's SVP of Product advises a hybrid model: one metric aligned with value (like seats) and another aligned with cost (like token usage) to ensure profitability.

Many AI coding agents are unprofitable because their business model is broken. They charge a fixed subscription fee but pay variable, per-token costs for model inference. This means their most engaged power users, who should be their best customers, are actually their biggest cost centers, leading to negative gross margins.

As more companies integrate AI, their costs are tied to variable usage (e.g., tokens, inference). This is causing a profound, economy-wide transformation away from predictable seat-based subscriptions towards more dynamic usage-based models to align costs with revenue.

The ARR/SaaS model, built on predictable human usage, is failing. AI agents can consume resources worth thousands of dollars for a low subscription fee, breaking the unit economics. This forces a shift to metered, consumption-based pricing similar to utilities like electricity.

Standard SaaS pricing fails for agentic products because high usage becomes a cost center. Avoid the trap of profiting from non-use. Instead, implement a hybrid model with a fixed base and usage-based overages, or, ideally, tie pricing directly to measurable outcomes generated by the AI.

Law firms currently benefit from subsidized, 'all-you-can-eat' AI pricing. As providers shift to consumption-based token pricing, the true, variable cost will emerge. This will likely cause 'sticker shock' and force a recalculation of AI's actual economic benefit.

Companies like coding assistant Cursor reportedly face negative gross margins because their flat-rate, per-seat pricing fails to cover the high compute costs from agentic tools that generate many tokens. This creates an unsustainable business model where growth exacerbates losses.