Product sense isn’t talent. It’s a gym routine. Do reps, gain instincts.
I’ll teach you in 3 minutes what took me 8 yrs.
The system: regular feature analysis with prediction tracking.
1. Pick one feature from any successful app.
2. Document your strategic hypothesis.
3. Predict their next observable move.
4. Verify in 4 weeks through public signals.
Example: Spotify wrapped feature →
Hypothesis:
"Spotify built Wrapped to drive social sharing during December when music discovery peaks and holiday playlist creation happens."
Strategic intent: viral growth through UGC, retention spike, brand visibility.
Business model tie: free user acquisition without paid spend.
Prediction (what you write today):
"They'll expand Wrapped beyond music - add podcasts, maybe audiobooks. Launch earlier each year. Create mid-year version to test retention impact."
Observable signals to check: app updates, blog announcements, social media buzz.
Four weeks later - verification:
Check app changelog: Did podcasts appear in Wrapped?
Check Spotify blog: Any mid-year campaign announced?
Check App Store: Feature updates mention Wrapped expansion?
Check social: User screenshots showing new categories?
Your prediction accuracy compounds.
The six exercises that build pattern recognition:
1. Strategic teardowns: why THIS feature, not alternatives?
2. Next move predictions: what ships in 30-60-90 days?
3. Business model mapping: how does this monetize?
4. Competitive response: what will rivals copy?
5. User psychology: what behavior change do they want?
6. Failure prediction: what could kill adoption?
Real patterns you'll recognize after 30 reps:
- When companies add social features, they're fighting retention decline.
- When they remove features fast, instrumentation showed zero usage.
- When they A/B test UI changes slowly, there's low confidence in the hypothesis.
- When they ship fast without testing, they're responding to a competitive threat.
The typical pattern:
PMs wait years for intuition to "click."
Then they document 30 strategic hypotheses.
Track predictions against observable outcomes.
Pattern recognition accuracy doubles by day 90.
The verification sources (all public):
- App Store changelog: "What's New" section shows iterations
- Product blogs: feature announcements, strategic updates
- App rankings: trending features drive ranking spikes
- Review sentiment: users call out what works/breaks
- Press releases: major strategic pivots
- Version history: expansion vs pullback patterns
Start today:
1. Pick Notion, Linear, or Figma.
2. Find one feature launched in the last 60 days.
3. Write why you think they built it.
4. Predict what they'll ship next.
5. Set calendar reminder for 4 weeks.
30 documented predictions = noticeable pattern recognition.
90 days = your prediction accuracy doubles.
hit the next tweet for some useful templates: 👇