The only guarantee in startups is that nothing is guaranteed.

PitchBook analyzed startup failure rates by stage and by “consensus-level” — defined as companies in the top decile of valuation.

A few things stood out to me:

➡️ 𝗙𝗮𝗶𝗹𝘂𝗿𝗲 𝗿𝗮𝘁𝗲𝘀 𝗿𝗲𝗺𝗮𝗶𝗻 𝘀𝗶𝗴𝗻𝗶𝗳𝗶𝗰𝗮𝗻𝘁 𝗲𝘃𝗲𝗻 𝗮𝘁 𝗹𝗮𝘁𝗲𝗿 𝘀𝘁𝗮𝗴𝗲𝘀.

We expect high failure rates at the early stage, but often assume they drop sharply over time. Even so, a successful Series B still carries a 30–40% chance of failure down the line.

➡️ 𝗧𝗼𝗽-𝗱𝗲𝗰𝗶𝗹𝗲 𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻𝘀 𝗰𝗼𝗿𝗿𝗲𝗹𝗮𝘁𝗲 𝘄𝗶𝘁𝗵 𝗵𝗶𝗴𝗵𝗲𝗿 𝘀𝘂𝗰𝗰𝗲𝘀𝘀 𝗿𝗮𝘁𝗲𝘀.

At the early stage, top-decile companies are 20–30% more likely to succeed than the rest.

Are those premium valuations justified? Or are companies that raise the most at the highest prices simply advantaged? Likely, both!

➡️ 𝗡𝗮𝗺𝗶𝗻𝗴 𝗰𝗼𝗻𝘃𝗲𝗻𝘁𝗶𝗼𝗻𝘀 𝗰𝗼𝗻𝘁𝗶𝗻𝘂𝗲 𝘁𝗼 𝗯𝗹𝘂𝗿 𝘁𝗵𝗲 𝗽𝗶𝗰𝘁𝘂𝗿𝗲.

For example, the top-decile Seed rounds in 2025 are priced around $40M pre, nearly 3x the stage median of $15M and equal to the median Series A.

Outstanding question for me: What’s happening at Series D & E+?

Maybe this explains why everyone piles into those consensus deals.