AI Pricing Paradox: Everyone Wants Your AI, But No One Wants to Pay

Got story updates? Submit your updates here. ›

The article discusses the paradox that while everyone wants to use AI, no one wants to pay for it. The author explains how hyperscalers subsidized AI to win market share, leading to seat-based pricing and a race to the bottom. As AI usage increases, companies are seeing their margins erode due to rising infrastructure costs. The article outlines four stages of AI pricing - the Trap, the Hedge, the Bet, and the Summit - and argues that the companies that figure out outcome-based pricing will be the next wave of AI winners, while those stuck in the Trap will be left behind.

Why it matters

The pricing of AI is a critical issue for companies as they struggle to balance the value their AI provides with the rising costs of the infrastructure to support it. Getting the pricing model right will be a key competitive advantage, as it will allow companies to capture the value of their AI investments and maintain healthy margins.

The details

The article explains that traditional SaaS pricing, where usage goes up but margins stay flat, doesn't work for AI, where increased usage leads to decreased margins due to rising infrastructure costs. Companies that have been giving away AI for free to win deals are now facing tough conversations with procurement teams as those pilots come up for renewal. The author argues that enterprises will pay for AI, but they want to pay for outcomes, not features.

• In 2024 and 2025, companies gave away AI in pilots to win deals.
• Those pilots are now hitting renewal cycles in the current quarter.

What’s next

The author recommends that companies start by adding an outcome bonus to their existing pricing, even if it's just 10%, to begin the measurement conversation and escape the 'Trap' of seat-based pricing. Once in the 'Hedge' stage, companies should tie pricing to a metric their customer already tracks, rather than inventing a new measurement.