Making the Decision (70% Rule)
"It's not about making the right decision, it's about making the decision." - Anneka Gupta (as remembered by colleague Rachel Wolan)
What It Is
Making the Decision is a framework for overcoming analysis paralysis by recognizing that you learn more after committing to a decision than you do by continuing to gather information beforehand. When you have roughly 70% confidence, that's enough to move forward—the remaining 30% will be refined through iteration.
The key insight: not deciding is itself a decision, and usually the worst one. You remain stuck in the hypothetical and gain no new high-quality information.
How It Works
The Core Principle:
- Before commitment: You operate on imprecise, hypothetical information
- After commitment: You get high-fidelity, real-world feedback
- The learning accelerates dramatically once you commit
The 70% Threshold:
- If you're ~70% confident, that's enough to decide
- You can iterate on the remaining 20-30% as you learn
- If you don't commit, you don't get new information to iterate on
The Learning Gap:
- Analysis gives you hypothetical insights
- Action gives you real insights
- Real insights are almost always more valuable than hypothetical ones
How to Apply It
Assess your confidence - Are you at roughly 70% confidence? If so, stop gathering more information and decide.
Document your hypothesis - Write down what you believe and why. This isn't just for accountability—it's so you can learn when reality differs from expectations.
Commit publicly - Making a decision real by sharing it with others increases learning velocity.
Plan for iteration - Build in checkpoints to assess whether your hypothesis was correct and adjust accordingly.
Expect partial waste - Accept that you might throw away 20% of work. This is better than making no decision at all.
Creating a Culture of Decisive Action
To help teams embrace this framework, Anneka recommends:
State hypotheses explicitly - When making decisions, be clear: "Our hypothesis is that [X segment] will pay for this because [Y reason]. Here's our evidence, and here's what we don't know."
Reward learning, not outcomes - If a decision didn't work out but the team learned something valuable, celebrate that learning.
Review hypotheses, not decisions - When something fails, go back to the original hypothesis. What did you learn that you couldn't have known before?
Make bad-decision tolerance explicit - Tell your team: "It's okay if we make bad decisions, as long as we learn from them and get better for next time."
When to Use It
- When you find yourself asking "If I just had one more data point..."
- When decisions keep getting delayed for more analysis
- When you're operating on imprecise information (which is always in product)
- When the cost of delay exceeds the cost of being 30% wrong
- When you need to build a culture of speed and learning
Source
- Guest: Anneka Gupta
- Episode: "Becoming more strategic, navigating difficult colleagues, founder mode, more"
- Key Discussion: (00:30:42) - Anneka explains why making decisions at 70% confidence is better than waiting
- YouTube: Watch on YouTube
Related Frameworks
See also: Thinking in Bets by Annie Duke (pending extraction)