analytical-thinking

Analytical Thinking Skill

Purpose

Analytical thinking for PMs is the structured ability to define metrics, build business cases, set goals, and make tradeoffs. This skill provides frameworks for each, based on Ben Erez's PM analytical thinking methodology.

Ecosystem Analysis

Why Ecosystem First

Products don't serve one user — they serve an ecosystem of players. A marketplace serves buyers AND sellers. A content platform serves creators AND consumers. A SaaS tool serves end users AND admins AND buyers.

Your metrics must capture value creation for ALL key players, not just the most visible one.

How to Map an Ecosystem

For each player (3-5 max):

1. Identify the player type (end users, creators, admins, buyers, advertisers)
2. State the primary value proposition to them
3. List the key actions they must take to realize value
4. Define metrics to track those actions

Metric format:

• # of [players] who [action] per [day/week/month] (measures reach)
• # of [actions] performed per [day/week/month] (measures volume)
• Avg [actions] per [player] per [day/week/month] (measures depth)

Critical: Always define what "active" means. Not "active users" but "users who completed [specific action] within [time period]."

North Star Metric Design

Selection Criteria

Your NSM must meet ALL of these:

Criteria	Test
Multi-player value	Growth means both users AND business win
No natural ceiling	Can grow continuously as product succeeds
Raw count preferred	Total count rather than percentage
Per time period	Per week/month, not cumulative
Actionable	Team can influence through product changes
Understandable	Every team member can explain it

NSM Examples by Product Type

Product Type	Good NSM	Bad NSM
Marketplace	Weekly transactions completed	GMV (cumulative, masks problems)
Content platform	Weekly content pieces with >1 engagement	Total uploads (ignores quality)
SaaS tool	Weekly active teams with >3 members using	MRR (lagging, not actionable)
Social network	Weekly connections with bidirectional interaction	Registered users (vanity)

Guardrail Design

For every NSM drawback, create a guardrail:

Pattern: NSM is a count → Guardrail is a rate/percentage

NSM Drawback	Guardrail Pattern
Could grow by adding low-quality items	% of items meeting quality threshold
Doesn't capture monetization	Revenue per active user
Could grow by adding unprofitable users	Unit economics (LTV/CAC)
Doesn't capture satisfaction	NPS or satisfaction score

Rule: If guardrail drops below threshold, NSM growth is unhealthy.

Time Period Selection

Factor	Daily	Weekly	Monthly
Consumer social	Often best	Good	Too slow
B2B SaaS	Too noisy	Often best	Good for enterprise
Marketplace	Good	Often best	Too slow
Transactional	Context-dependent	Good	Often best

Choose based on: natural usage cadence, signal-to-noise ratio, ability to act on data.

Metrics Hierarchy

North Star Metric
├── L1: Direct drivers (team can influence)
│   ├── L2: Input metrics (individual can influence)
│   └── L2: Input metrics
├── L1: Direct driver
│   ├── L2: Input metric
│   └── L2: Input metric
└── Guardrails (quality/safety checks)

L1 metrics directly move the NSM. If L1 improves, NSM improves. L2 metrics are inputs to L1. These are what individual contributors optimize.

Goal Setting Framework

The Altitude Shift

NSM is 50,000 feet. Team goals are ground level. The altitude shift connects them:

1. Select ONE ecosystem player to focus on (with rationale)
2. Map their complete journey to NSM contribution
3. Identify conversion rates between journey steps
4. Find the biggest drop-off — that's where goals live
5. Set 3 candidate goals as conversion rate improvements
6. Score each on Impact (how much it moves NSM) × Ability to Influence (can the team actually move this)
7. Select ONE primary goal

Goal Guidelines

• Goals = conversion rate improvements between journey steps
• Do NOT set averages (they hide segment differences)
• Define directionally first ("increase X"), then set specific targets
• Always pair with anti-goals (what shouldn't get worse)

Tradeoff Framework

Tradeoff Types

Type	Tension
Breadth vs Depth	Many features vs few excellent ones
Speed vs Quality	Ship fast vs ship right
Short-term vs Long-term	Quick win vs lasting investment
User A vs User B	Optimize for one segment vs another
Build vs Buy	In-house vs third party
Simplicity vs Power	Easy to use vs feature-rich
Revenue vs Growth	Monetize vs expand the base

Analysis Structure

1. Name the common goal both options serve
2. Frame each option with exactly 2 pros and 2 cons
3. Assess evidence for each claim (data vs assumption)
4. Check strategic alignment (mission, NSM, segments, pillars)
5. Make a decisive recommendation with rationale
6. State what would change your mind

Decision Heuristics

When analysis doesn't clearly favor one option:

• Choose the most reversible option — you can always change later
• Choose the option that generates learning — data > guessing
• Choose the option closest to your strategy — consistency compounds
• When in doubt, do less — simplicity is a feature