Flat AI Spend Is Not Flat AI Demand: What Ramp's Data Means for Your Pricing

Last updated: November 2026

Key takeaways

• Per TechCrunch's report on Ramp data, token spend grew 42% in July, then 1.5% across August and September combined, then 8.9% from September 23 to October 23.
• Token volumes kept rising during the stall, so the spend plateau is a price story, not a demand story.
• OpenAI halved API prices for GPT-6 Sol and GPT-6 Luna in September, and Anthropic cut prices on Fable 5.1 and Opus 5.5.
• The spread between median and heavy users is huge: $12 to $15 per employee per month at the median versus $600 to $1,000 in the top 10% of firms.
• If you price per seat, your heaviest customers are worth 40 to 80x your median one in COGS.

What did Ramp's data actually show?

According to TechCrunch's report on Ramp's data, business token spend rose 42% in July, only 1.5% across August and September combined, then 8.9% in the month to October 23. Token volumes kept rising throughout. The top 1% of firms spent just under $10,000 per employee per month at their summer peak, and dipped slightly after.

The headline reads as a demand scare right before Anthropic's IPO. We think that is the wrong reading.

Why flat spend does not mean flat demand

Spend equals volume times price. If prices fall, spend can stay flat while volume grows. September is when OpenAI halved prices for GPT-6 Sol and Luna and Anthropic cut Fable 5.1 and Opus 5.5, a move Anthropic said reduced its revenue growth by about $600 million, per the same report.

| Price change | Volume growth needed to keep spend flat | |---|---| | -20% | +25% | | -33% | +50% | | -50% | +100% |

Illustrative arithmetic: a 50% price cut means customers must double their token volume just to hold the same invoice. Flat spend with rising volume is what a price war looks like from the buyer side. This echoes the Jevons paradox for AI demand we covered earlier.

What does the median-versus-top gap mean for seat pricing?

The median firm spends $12 to $15 per employee per month on tokens; the top 10% spend $600 to $1,000. That is roughly 40 to 80x. Say you charge $30 per seat per month for an AI feature. A median user costs you $12 to $15 in tokens, which is a 50% to 60% COGS ratio before anything else, and a top-decile user costs $600 or more, which is negative margin at any seat price. Flat seat pricing silently subsidizes your heaviest users. We compared the options in flat vs usage-based AI pricing.

What should SaaS founders do now?

1. 1Track tokens per customer, not only dollars. Dollars move with provider price cuts; tokens show real adoption.
2. 2Pass price cuts through carefully. If your provider cuts rates 50%, your COGS drops, and so does your competitors' price. Decide your repricing move in advance.
3. 3Add usage tiers or credits for the top decile, so heavy users pay in proportion to cost.
4. 4Re-run your forecast at the new rates. Use the AI chatbot or AI agent use case pages as a starting point.