💡

InsightHunt

Hunt the Insights

C

Crystal Widjaja

Chief Product Officer at Kumu / Former SVP of Growth at Gojek

Kumu / Gojek (GoTo)

📈 Growth & Metrics (1)🎯 Product Strategy (1)⚡ Execution (1)

Key Takeaways

  • 1.Run experiments even with small sample sizes (N=30); the precision changes, but the directional trend rarely does.
  • 2.For free consumer products, aim for 60% Week 1 retention that flattens immediately; if early adopters drop off, mass market users definitely will.
  • 3.Stop treating 'retention' as a goal; it is a result. Focus on the 'setup moment'—the step immediately preceding the conversion event.
  • 4.Most analytics fail because they track observations (events), not context. You must track the state of the world (properties) at the time of the event to derive insight.
  • 5.Utilize 'Physics vs. Levers' thinking: Understand the immutable constraints of your market (Physics) before trying to optimize your growth loops (Levers).
  • 6.When hiring growth roles, prioritize statistical intuition and 'first principles' thinking over tool proficiency.
  • 7.Use 'Wizard of Oz' testing (e.g., manual WhatsApp groups, static screenshots) to validate complex features like subscriptions before engineering builds them.

Methodologies(3)

📈 Growth & Metrics

A data tracking philosophy that shifts focus from counting events (observations) to capturing the state of the user and environment (context) at the moment of the event. This transforms data from a scorecard into a diagnostic tool.

Core Principles

  • 1.Step 1: Differentiate Observation vs. Insight. An observation is 'User booked a ride'. An insight explains *why* based on context.
  • 2.Step 2: Map the User's Reality. When defining an event, list every environmental variable visible to the user (e.g., price, weather, number of items on screen).
  • 3.Step 3: Instrument Heavy Properties. Do not just track `map_loaded`. Track `map_loaded` with properties: `num_drivers_visible`, `surge_pricing_active`, `weather_condition`, `pickup_distance`.
  • +1 more...

"If I see my girlfriend hanging out with a guy I don't know, that is an observed fact... The insight is, 'I am paranoid.' The insight provides value when you have this 'why' answered."

#context-first#instrumentation#growth
View Deep Dive →
🎯 Product Strategy

A strategic framework that separates the immutable constraints of the business environment (Physics) from the variables you can control and optimize (Levers). Growth comes from fitting levers into the existing physics, not trying to change physics.

Core Principles

  • 1.Step 1: Define the Physics. Audit the immutable realities: Market (who are they?), Product (what is it?), Model (how do we charge?), and Channel (where do they live?).
  • 2.Step 2: Identify Existing Assets. Look for underutilized assets within that physics (e.g., captive attention, existing supply).
  • 3.Step 3: Design the Lever. Create a mechanism that utilizes the asset to drive the goal without changing the core user behavior.
  • +1 more...

"We are not wizards. It's very hard to move the physics of a universe... Start with what currently works and exists."

#physics#levers#growth
View Deep Dive →
⚡ Execution

A methodology for validating demand and user experience by manually simulating the product backend. It prioritizes the 'front-stage' user experience over 'back-stage' scalability to get immediate data.

Core Principles

  • 1.Step 1: Define the Ideal Experience. What does the user *see* and *feel* if this feature works perfectly?
  • 2.Step 2: Hack the Interface. Use existing tools (WhatsApp, Typeform, Screenshots via In-App Message) to present the offer.
  • 3.Step 3: Manual Execution. Use humans (interns, founders) to manually perform the backend logic (e.g., checking databases, sending vouchers).
  • +2 more...

"If you don't have a tested hypothesis... honestly that idea is pretty useless. Testing the actual experience and seeing how people respond to it, that's the best possible data."

#'wizard#validation#protocol
View Deep Dive →