
The quality of your deal flow determines the quality of your portfolio. Yet most VC firms manage their pipeline with a chaotic mix of spreadsheets, email threads, and someone's memory. If you're the associate responsible for keeping deal flow organised, here's how to build a system that actually works.
Before you can manage deal flow, you need to understand where it comes from. Track the source of every deal for three months, and you'll likely find:
The deals from proactive sourcing typically convert at 3-5x the rate of inbound. That's where you should be spending more time.
Every deal should move through clearly defined stages:
Tag every deal with its current stage and update weekly. Your partners should be able to see the full pipeline at a glance.
You can't give equal attention to everything. Use a simple 2x2 matrix: conviction (high/low) vs. urgency (competitive process/no timeline pressure). Deals in the high-conviction, high-urgency quadrant get your time first.
For quick conviction signals on new deals, run them through Predict Ventures. The benchmarking data helps you rapidly assess whether a startup's metrics are genuinely impressive or merely average for its stage.
Just because you pass today doesn't mean the company won't be interesting in 18 months. Set follow-up reminders for promising companies that were too early or had fixable weaknesses. Some of the best investments come from the "not yet" pile.
Block 30 minutes every Friday to review your pipeline:
Share a weekly pipeline summary with your partners. It keeps them informed and shows you're on top of things.
Whether you use Affinity, Attio, or a spreadsheet, the system only works if the data is clean. Make it a habit to update deal status, add notes after every call, and tag deals by sector, stage, and source. Your future self—and your colleagues—will thank you.
Track these metrics monthly:
These numbers tell you where your process is breaking down and where to invest your energy.
Stop relying on gut feel. Predict Ventures benchmarks every startup against 15,000+ data points and 50 years of exit history to give you a quantitative edge.