Pacific Northwest startups most likely to have a successful exit, according to PitchBook

Abstraction: ML tool predicting VC-backed startup exit likelihood and type

Key points:

• PitchBook launched the VC Exit Predictor, a machine learning tool trained on 120,000+ VC-backed companies to forecast exit likelihood and type (acquisition, IPO, self-sufficient, or shutdown)
• Model is based on 46,000 observations with known exit outcomes; tested on 11,000+ held-out cases with 75% accuracy
• Outputs an "opportunity score" percentile and projected exit path; does not use founder personal characteristics (education, gender) to minimize bias
• US total exit deal value fell to $71.4B in 2022 — down 90%+ from prior year — the first time since 2016 below $100B
• CB Insights' similar Mosaic Score faced criticism in 2021 for incorporating founder educational background and network quality, potentially exacerbating investment bias

Connections: Pitchbook · Cb Insights · Machine Learning · Venture Capital · Predictive Analytics

Source: https://www.geekwire.com/2023/pacific-northwest-startups-most-likely-to-have-a-successful-exit-according-to-pitchbook/