North Star Metric: Tìm và Theo Đuổi Metric Quan Trọng Nhất

TL;DR

North Star Metric (NSM) = Single metric best capturing core customer value của product
Examples: Airbnb (Nights booked), Facebook (DAU), Netflix (Hours watched), Slack (Messages sent)
Characteristics: Reflects customer value, Predictive of revenue, Actionable, Simple
How to find: Map customer journey → Identify "aha moment" → Define metric around it
Supporting metrics: Input metrics driving NSM (acquisition, activation, retention, etc.)
Avoid: Vanity metrics (registered users ≠ active value), Too many NSMs (lose focus)
Tracking: Dashboard, weekly reviews, tie to OKRs
Impact: Aligns entire company to one goal, clarity in decision-making

Giới Thiệu: The Metric Overload Problem

Scenario thường gặp:

Weekly metrics review meeting:

CEO: "How are we doing?"

PM: "Great! Registered users up 50%"
CMO: "Page views up 30%"
CTO: "Uptime 99.9%"
CFO: "Revenue flat... 🤔"

CEO: "We're growing all these metrics but revenue isn't moving. Which metric should we focus on?"

Team: *Confusion*

Vấn đề: Too many metrics → No clarity on what truly matters.

The reality:

Average company tracks 50-100 metrics
Executives look at 10-20 in weekly meetings
But only 1-3 metrics truly drive the business

Result:

Teams optimize different metrics (mis-aligned)
Vanity metrics look good but don't drive value
Decision paralysis ("This metric up, that metric down, what do we do?")

Solution: North Star Metric - The ONE metric that matters most.

What is North Star Metric?

Definition: Single metric that best captures the core value you deliver to customers.

Metaphor: North Star = Guiding star for sailors

Always points North (true direction)
Everything else is navigation details

In business:

NSM = True value your product delivers
All other metrics = How to get there

Characteristics of Great North Star Metric

1. Reflects Customer Value

Good NSM measures value TO CUSTOMERS, not just company:

❌ Bad: Revenue
- Company metric, not customer metric
- Doesn't tell you if customers happy

✅ Good: Nights booked (Airbnb)
- Customer getting value: place to stay
- More nights = more customer value
- Revenue follows naturally

2. Predictive of Revenue

NSM should be leading indicator of revenue (not lagging):

Lagging: Revenue (know it after fact)
Leading: Active users → Will convert to revenue

If NSM up → Revenue will follow (eventually)
If NSM flat → Revenue growth will stall

3. Actionable by Team

Team can directly influence NSM:

❌ Not actionable: "Market size"
- Can't control market size

✅ Actionable: "Weekly active users"
- Can run campaigns to activate users
- Can improve product to retain users

4. Simple to Understand

Everyone in company can explain NSM:

Test: Ask any employee "What's our North Star?"
✅ Good: Everyone says same thing
❌ Bad: Different answers, confusion

5. Measurable & Trackable

Can track NSM daily/weekly:

✅ Measurable: "Messages sent per week"
❌ Hard to measure: "Customer happiness" (subjective, survey-based)

Examples: North Star Metrics of Successful Companies

Airbnb: Nights Booked

Why not "Users" or "Listings"?

Users registered: Vanity metric (may never book) Listings created: Supply metric (not customer value)

Nights booked: Perfect NSM

Guest value: Got place to stay
Host value: Earned income
Airbnb: Revenue = nights booked × take rate

All decisions align:

Product: Make booking easier → More nights
Marketing: Attract guests → More nights
Operations: Support hosts → More nights

Facebook: Daily Active Users (DAU)

Why not "Registered users"?

Facebook has 3 billion registered users, but:

Many inactive (created account, never returned)
Registered users ≠ engaged users

DAU: Active = Real value

User getting value: Connecting with friends
Facebook: More DAU → More ad impressions → Revenue

Supporting metrics (drive DAU):

New user activation (7 friends in 10 days)
Engagement (posts, likes, comments)
Retention (% coming back next day)

Netflix: Hours Watched

Why not "Subscribers"?

Subscribers: Important, but lagging

Subscriber who watches 0 hours → Will churn
Subscriber who watches 50 hours/month → Sticky, won't churn

Hours watched: Leading indicator

More hours → Higher retention → Higher LTV
Content quality → More hours watched

Actions:

Invest in content that drives hours watched
Improve recommendation algorithm → More hours
Optimize UI for binge-watching

Slack: Messages Sent

Why not "Teams signed up"?

Team signs up but sends 10 messages/week → Not getting value, will churn

Messages sent: Collaboration happening

Team sending 1,000+ messages/week → Deeply engaged
Sticky, high retention

2,000 messages = Magic number

Slack discovered: Teams sending 2,000+ messages have 93% retention
Activation goal: Get teams to 2,000 messages ASAP

E-commerce: Orders Placed

Why not "Revenue"?

Revenue = orders × average order value (AOV)

Revenue up could mean: More orders OR higher prices
If price increase → Revenue up but orders down → Fewer customers (bad long-term)

Orders placed: Customer transactions

More orders = more customers getting value
Revenue follows

Supporting metrics:

Conversion rate (visitors → buyers)
Repeat purchase rate
AOV (optimize, but not at cost of orders)

How to Find Your North Star Metric

Step 1: Map Customer Journey

Example: E-learning app

Customer journey:
1. Discover app (Google, ads)
2. Sign up
3. Browse courses
4. Start first course
5. Complete lessons
6. Earn certificate
7. Apply knowledge (get promoted, new job)

→ Where is "aha moment"? (When customer realizes value)

Aha moment: Complete first course, earn certificate

Customer: "Wow, I learned something valuable!"
Likely to come back for more courses

Step 2: Identify "Aha Moment"

"Aha moment" = When customer experiences core value

Examples:

Product	Aha Moment
Dropbox	First file synced across devices
Uber	First successful ride completed
Tinder	First match
Duolingo	First lesson completed
Zoom	First successful video call

How to find:

Talk to customers: "When did you realize this product is valuable?"
Cohort analysis: Users who did X have 5x higher retention
Survey: NPS score spike after which action?

Step 3: Define Metric Around Aha Moment

E-learning app example:

Aha moment: Complete first course

NSM candidates:
1. Courses completed (too narrow, one-time)
2. Weekly active learners (vague, what's "active"?)
3. Learning hours per week (best!)

→ NSM: Learning hours per week

Why "Learning hours"?

Reflects value: More learning = more knowledge
Recurring: Weekly metric (not one-time)
Predictive: More hours → Higher retention → More subscriptions

Step 4: Validate with Data

Test: Does proposed NSM correlate with business outcomes?

-- Check correlation: Learning hours vs Retention
WITH user_cohorts AS (
    SELECT
        user_id,
        AVG(learning_hours_per_week) as avg_hours,
        CASE
            WHEN last_active_date >= CURRENT_DATE - 30 THEN 'Retained'
            ELSE 'Churned'
        END as retention_status
    FROM user_activity
    GROUP BY 1
)

SELECT
    retention_status,
    AVG(avg_hours) as avg_learning_hours
FROM user_cohorts
GROUP BY 1;

-- Result:
-- Retained: 8.5 hours/week
-- Churned: 2.1 hours/week

→ Strong correlation! NSM validated.

Step 5: Workshop with Team

Facilitated session (2-3 hours):

Participants: Product, Engineering, Marketing, Exec team

Agenda:

Review customer journey (30 min)
- Map stages
- Identify aha moment
Brainstorm NSM candidates (30 min)
- Each person proposes 1-2 metrics
- Write on whiteboard
Evaluate against criteria (45 min)
- Customer value? ✅ / ❌
- Predictive of revenue? ✅ / ❌
- Actionable? ✅ / ❌
- Simple? ✅ / ❌
- Measurable? ✅ / ❌
Vote & decide (15 min)
- Narrow to top 2-3
- Final decision (CEO / PM)
Define supporting metrics (30 min)
- Input metrics that drive NSM
- How to track

Output: Consensus on NSM, alignment across teams.

Supporting Metrics: The NSM Tree

NSM alone is not enough. Need input metrics (drivers).

Example: Slack's Metric Tree

North Star Metric: Messages Sent per Week
                    ↑
        ┌───────────┼───────────┐
        │           │           │
   Acquisition  Activation  Retention
        │           │           │
    New teams   Time to     Daily Active
    signed up   2,000 msgs  Teams (DAT)

How they connect:

More teams signed up → More activation → More retained teams → More messages sent

Actions per metric:

Acquisition: Marketing campaigns, SEO, referrals
Activation: Onboarding flow, invite teammates, send first messages
Retention: Feature adoption, notifications, integrations

Pirate Metrics (AARRR Framework)

Popular framework for SaaS products:

Acquisition → Activation → Retention → Revenue → Referral
     ↓            ↓           ↓          ↓         ↓
   Visitors   Sign-ups    Day 7      Paying    Invites
   to site    complete    return     customers sent
              onboarding

NSM typically at Retention or Revenue stage:

Retention-focused NSM: WAU, DAU
Revenue-focused NSM: Transactions, subscriptions

Avoiding Vanity Metrics

Vanity metrics: Look impressive but don't drive business value.

Examples of Vanity Metrics

1. Registered Users

Metric: 1 million registered users!
Reality:
- 900K inactive (signed up, never returned)
- 100K active

→ Misleading

Better: Monthly Active Users (MAU)

2. Page Views

Metric: 10 million page views/month
Reality:
- 90% bounce rate (leave after 1 page)
- No engagement, no conversions

→ Doesn't reflect value

Better: Time on site, Pages per session, Conversion rate

3. App Downloads

Metric: 500K downloads
Reality:
- 80% delete app after 1 use
- 100K active users

→ Downloads ≠ engagement

Better: DAU, Retention rate

4. Social Media Followers

Metric: 100K Twitter followers
Reality:
- 5% engagement rate
- Most followers inactive/bots

→ Followers ≠ influence

Better: Engagement rate, Click-through rate

How to Spot Vanity Metrics

Ask:

Can I act on this metric?
- If no → Vanity
Does it predict business outcomes?
- Test correlation with revenue/retention
- If weak correlation → Vanity
Would customers care?
- "Registered users" = Company cares, customers don't
- "Messages sent" = Customers care (getting value)

Tracking Your North Star Metric

1. Build NSM Dashboard

Components:

North Star Dashboard:

┌─────────────────────────────────────┐
│ North Star Metric: Learning Hours  │
│                                     │
│  This week: 125,000 hours          │
│  Last week: 120,000 hours          │
│  Change: +4.2% ↗                   │
│                                     │
│  [Line graph: 12-week trend]       │
└─────────────────────────────────────┘

Supporting Metrics:
┌──────────────┬──────────────┬──────────────┐
│ New Users    │ Activation   │ Retention    │
│ 5,000        │ 65%          │ 55%          │
│ +10% ↗      │ +2% ↗       │ -1% ↘       │
└──────────────┴──────────────┴──────────────┘

Tools: Looker, Tableau, Metabase

Pin dashboard to TV in office
Share in Slack channel weekly

2. Weekly NSM Review

Meeting (30 mins, every Monday):

Attendees: Product, Eng, Marketing leads

Agenda:

NSM performance (5 min)
- Up/down/flat?
- Why?
Dive into drivers (15 min)
- Which input metrics moved?
- Acquisition up but retention down → Net neutral
Action items (10 min)
- What to do this week to move NSM?
- Assign owners

Example:

NSM: -2% this week (Learning hours down)

Why?
- Retention: -5% (Spring break, students on vacation)
- Activation: Flat

Action:
- Marketing: Launch "learn during break" campaign
- Product: No action (seasonal dip expected)

Next week target: Back to baseline when school resumes

3. Tie to OKRs

OKR (Objectives & Key Results):

Objective: Become #1 learning app

Key Results (Quarterly):
1. Increase Learning Hours/week from 120K → 150K (+25%)
2. Improve activation rate from 65% → 75%
3. Reduce churn from 5% → 3% monthly

→ All KRs tie to NSM (Learning Hours)

Benefits:

Company-wide alignment
Clear success criteria
Motivation (progress visible)

Common Mistakes

Mistake 1: Too Many North Stars

❌ Bad:
"Our North Stars are:
- Monthly Active Users
- Revenue
- Customer Satisfaction
- Net Promoter Score"

→ If everything is priority, nothing is priority

Fix: One North Star. Other metrics are supporting or guardrails.

Mistake 2: Picking Revenue as NSM

❌ Revenue as NSM
- Lagging indicator (already happened)
- Doesn't tell you WHY revenue moved
- Can game (raise prices → revenue up but customers down)

Better: Leading indicator that predicts revenue

Active users → Will generate revenue
Transactions → Revenue follows

Exception: Early-stage startups can use revenue (need to survive first)

Mistake 3: Ignoring Trade-offs

Example: Video app
NSM: Hours watched

Team optimizes for hours → Auto-play next video aggressively

Result:
- Hours watched ↗
- User satisfaction ↘ (annoyed by aggressive auto-play)

→ NSM up, but experience degraded

Fix: Guardrail metrics

NSM: Hours watched
Guardrails: User satisfaction score (must stay >4/5), Churn rate (must stay <5%)

Rule: Increase NSM WITHOUT hurting guardrails.

Mistake 4: Set & Forget

Company defines NSM in 2020: "Monthly Active Users"

2023: Product evolved, MAU not right metric anymore

But nobody revisited → Team optimizing wrong metric

Fix: Review NSM annually

Is it still reflecting customer value?
Business evolved? NSM should too

Case Study: Vietnamese Food Delivery App

Background

Company: Food delivery app, Series A Challenge: Tracking 30+ metrics, no focus

Metrics tracked:

App downloads
Registered users
Daily active users
Orders per day
GMV (Gross Merchandise Value)
Restaurants onboarded
Delivery time
Customer satisfaction
...

Problem: Teams optimizing different metrics

Marketing: Focused on downloads
Product: Focused on DAU
Ops: Focused on delivery time

Result: Misalignment, slow progress.

Finding North Star Metric (Workshop)

Customer journey mapped:

1. Download app
2. Browse restaurants
3. Place first order
4. Food delivered
5. Rate experience
6. Place repeat order ← Aha moment!

Insight: Customer realizes value when they order AGAIN (trust established).

NSM candidates:

Total orders/week (chosen!)
DAU (too high-level)
GMV (company metric, not customer metric)

Decision: NSM = Orders per Week

Reflects customer value (getting food delivered)
Predictive of revenue (GMV = orders × AOV)
Actionable (team can drive orders)

Supporting Metrics Defined

North Star: Orders per Week
            ↑
    ┌───────┼───────┐
    │       │       │
New users  Repeat   Avg orders
acquiring  rate     per user

Targets (Quarterly OKR):

NSM: 50K orders/week → 75K (+50%)
New user acquisition: 10K/week → 15K
Repeat rate: 30% → 40%
Avg orders/user: 1.5/week → 2.0

Results (6 Months)

Before NSM focus:

30+ metrics tracked, no clarity
Marketing & Product mis-aligned
Orders: 50K/week (flat)

After NSM focus:

ONE metric everyone focused on: Orders/week
All decisions evaluated: "Will this increase orders?"
Product: Improved re-order flow (1-click reorder) → Repeat rate ↗
Marketing: Shifted budget to retention campaigns (not just acquisition)

After 6 months:

Orders: 50K → 90K/week (+80%!)
Repeat rate: 30% → 45%
GMV: 2x growth

CEO: "NSM aligned entire company. Best decision we made."

Kết Luận

Key Takeaways

✅ North Star Metric = ONE metric capturing core customer value ✅ Examples: Airbnb (Nights booked), Facebook (DAU), Netflix (Hours watched) ✅ Characteristics: Customer value, Predictive, Actionable, Simple, Measurable ✅ How to find: Map journey → Aha moment → Define metric → Validate ✅ Supporting metrics: Input drivers (acquisition, activation, retention) ✅ Avoid: Vanity metrics (registered users), Multiple NSMs, Revenue as NSM ✅ Track: Dashboard, Weekly reviews, OKRs ✅ Impact: Company alignment, Clarity, Faster progress

Recommendations

For Early-Stage Startups:

NSM = Revenue (survival first)
Once PMF → Switch to customer value metric

For Growth-Stage:

Define NSM via workshop (align team)
Track weekly, tie to OKRs
Review annually

For Enterprises:

Different NSMs per business unit OK
Consolidate to company-level NSM for exec team

Next Steps

Workshop Template (run with your team):

1. Map customer journey (30 min)
2. Identify aha moment (15 min)
3. Brainstorm NSM candidates (30 min)
4. Evaluate against criteria (45 min)
5. Vote & decide (15 min)
6. Define supporting metrics (30 min)

Total: 2.5 hours

Output: Your North Star Metric + supporting metrics

Carptech can facilitate your NSM workshop (virtual or in-person).

📞 Liên hệ Carptech: carptech.vn

Related Posts: