When headline valuations aren't what they seem

Summary

The AI funding boom has accelerated the use of two-tranche venture rounds, in which investors in the same deal pay different valuations—often rewarding early backers and lead investors—creating a gap between headline valuations reported in news and the actual prices paid, as documented in deals like AI-native IT startup Serval's December round where Sequoia's lowest entry valued the company at less than half its $1 billion headline figure.

Market Impact

The proliferation of tranched rounds reflects a power shift toward founders in an overheated AI funding market where lead VCs are capturing upward of two-thirds of rounds per Carta data, and demand is so intense that paying a premium for cap-table access becomes the only entry path. AI startups raised $255.5 billion globally in Q1 2026—surpassing the full-year 2025 total in a single quarter per PitchBook—intensifying competition for coveted deals. The practice obscures true valuation signals for the broader market and limited partners, as headline figures may overstate the consensus price by more than 100%, complicating valuation benchmarking and mark-to-market accounting across venture portfolios.

Why It Matters

Tranched rounds that detach headline valuations from actual prices paid distort the valuation signals that limited partners, later-stage investors, and the broader market rely on to benchmark AI startup worth, raising transparency concerns during the funding boom.

Key Points

• Two-tranche venture rounds let investors in the same funding round pay different valuations, with the AI boom enabling founders to reward early believers and lead investors through this structure
• In AI-native IT startup Serval's December round led by Sequoia at a $1 billion valuation, The Wall Street Journal reported Sequoia's lowest entry valued the company at less than $400 million, under half the headline figure
• AI startups raised $255.5 billion globally in Q1 2026, surpassing the full-year 2025 total in a single quarter according to PitchBook's Q1 2026 AI VC Trends report
• Lead VCs are taking upward of two-thirds of rounds per Carta data, pushing for better terms in return for higher-conviction, larger-check bets as rounds become heavily oversubscribed

Key Entities

Evidence

While letting investors in the same funding round at different valuations isn't a new invention, the AI boom has enabled some founders to play favorites in this way.

Supports: Confirms the tranched-round practice and its acceleration during the AI boom

Serval's December round, led by Sequoia at a $1 billion valuation, combined two classes of preferred shares priced at different valuations in the same deal. The Wall Street Journal reported that Sequoia's lowest Serva...

Supports: Grounds the gap between headline and actual valuations with the Serval example

AI startups raised $255.5 billion globally in the first quarter of 2026, surpassing the full-year 2025 total in a single quarter, according to PitchBook's Q1 2026 AI VC Trends report.

Supports: Documents the funding intensity driving competition for AI deal access