• Home
  • BVSSH
  • C4E
  • Playbooks
  • Frameworks
  • Good Reads
Search

What are you looking for?

Standard : Number of Experiments Run

Description

Number of Experiments Run measures how many structured experiments a team conducts in a given period (e.g. prototypes, A/B tests, concierge tests). It indicates how frequently teams are testing assumptions and learning.

Higher volume generally signals a culture of experimentation, provided each experiment is meaningful and tied to a hypothesis.

How to Use

What to Measure

  • Count of all discovery or growth experiments started and completed in a given period.
  • Include both successful and unsuccessful experiments.

Formula

Experiment Count = Total Experiments Conducted in Period

Example: Team runs 12 experiments in Q1 → Experiment Count = 12.

Instrumentation Tips

  • Maintain an experiment backlog or register.
  • Track start and end dates to measure cadence.
  • Capture experiment outcomes for later analysis.

Why It Matters

  • Learning velocity: Indicates how fast the team is generating insights.
  • Risk reduction: Validates ideas before large-scale investment.
  • Innovation culture: Reinforces data-driven decision-making.

Best Practices

  • Keep experiments small, fast, and cheap.
  • Tie each experiment to a clear hypothesis.
  • Document learnings and share them with the wider organisation.

Common Pitfalls

  • Counting vanity experiments that don't test real assumptions.
  • Not completing experiments or failing to extract learnings.
  • Over-investing in a single large test instead of iterating.

Signals of Success

  • Steady or increasing experiment volume per quarter.
  • Faster insight-to-action cycles.
  • Roadmap adjustments driven by experiment results.

Related Measures

  • [[CoE/Product/Measures/Discovery Effectiveness/Experiment Success Rate]]
  • [[Learning Velocity]]
  • [[A/B Test Coverage]]

Technical debt is like junk food - easy now, painful later.

Awesome Blogs
  • LinkedIn Engineering
  • Github Engineering
  • Uber Engineering
  • Code as Craft
  • Medium.engineering