Managing R&D is about making a portfolio of bets with deliberate risk levels, while protecting that investment from being consumed by day-to-day delivery. You cannot guarantee outcomes, so you manage the portfolio and the process.
How to think about it
R&D PORTFOLIO APPROACH
- Spread bets across risk: safe / moderate / moonshot
- Stage-gate: small experiments, double down on what works, kill the rest fast
- Time-box and set clear hypotheses for each bet
- Separate funding/teams so R&D is not raided for delivery
- Measure learning, not just immediate ROI
Treat R&D like venture investing: many small bets, fast kills, and occasional big wins, rather than expecting every project to pay off.
Concrete example
A CTO funds five small explorations per year with clear hypotheses and 8-week limits, kills three early, scales one into a product line, and explicitly accepts the failures as the cost of the one success.
