Customer Concentration
Customer concentration measures how much revenue depends on a few customers · AI labs often sit at 40-60% from top 10 vs SaaS at <20%.
Customer concentration measures how much revenue depends on a few customers · AI labs often sit at 40-60% from top 10 vs SaaS at <20%.
Basic
If a company's top 10 customers generate 60% of revenue, customer concentration is high · a risk factor. In classic SaaS, top-10 concentration of <20% signals product-market-fit across a broad customer base. AI labs trend high because enterprise deals are massive · one Azure/Google contract can equal thousands of SMB subscriptions. Regulators and investors scrutinize concentration in diligence.
Deep
OpenAI reportedly has Microsoft as >50% of revenue via revenue-share + Azure commitments. Anthropic has Amazon + Google as major customers post-Series E · concentration >40%. Mistral's enterprise tilt gives it 60%+ from top 10. High concentration = fragile (losing one customer hurts) but also validating (enterprise buyers staked their AI strategy). Low concentration = robust but usually means lower absolute revenue.
Expert
Concentration math: measure revenue from top N customers / total revenue. Typical cuts: top 1, top 5, top 10. SaaS diligence benchmarks: top 10 < 20%. AI: top 10 < 50% is typical. Public filings require disclosure of any customer >10% of revenue · OpenAI, Anthropic are private but leaks suggest Microsoft/Amazon both cross 10% for respective labs. Mitigation: diversifying across enterprise, API, and consumer channels · trend for frontier labs is to dilute concentration by scaling API + consumer.
Depending on why you're here
- ·Revenue share from top N customers
- ·SaaS: top 10 < 20% · AI: 40-60%
- ·Public filing threshold: any single customer >10%
- ·Not actionable at team level
- ·Signals partner risk when choosing an AI provider
- ·Diversify to avoid being a vendor's 10%+ customer
- ·Primary risk factor in AI-lab diligence
- ·High concentration = fragile but validating
- ·Watch trajectory · concentration should decline as revenue scales
- ·How much money comes from just a few customers
- ·Risky if too much · one bad deal could hurt the business
- ·AI companies tend to have more concentration than other tech
AI customer concentration is the biggest unhedged risk most frontier labs carry · watch how they diversify as API and consumer scale.