The Metrics & Milestones Map: A Stage-by-Stage Guide
TL;DR
- Most product teams default to the same dashboard : DAUs, MRR, churn, NPS.
- Every product moves through roughly five stages before it reaches scale.
- This is the full map. For each stage, here is what to measure, what milestone signals you've passed the stage, and what the leading indicator is that tells you before the lagging indicator confirms it.
- Tracking MRR before you have retention.
Most founders track metrics. The ones who raise and scale track the right metrics at the right stage.
📊 Read time: 14 minutes. Use time: every board meeting, every fundraise, every quarter you're deciding what to build next.
Why This Exists
Most product teams default to the same dashboard: DAUs, MRR, churn, NPS. These are real numbers. They are also almost always the wrong numbers to optimize at the wrong time.
A pre-product-market-fit team obsessing over churn is solving the wrong problem. A growth-stage team without a clear retention curve has no business running paid acquisition. The mistake is not measuring things. The mistake is measuring stage-agnostic metrics as if stage doesn't exist.
The teams that build something defensible treat metrics as stage-specific contracts. What you measure in the first 90 days is different from what you measure in month 18, which is different from what you measure heading into a Series A. This guide maps those contracts clearly, so you always know what signal you're actually looking for, and what milestone tells you it's time to shift gears.
How to Use This
- Find your current stage. Use the stage definitions in Section 1 to place yourself honestly. When in doubt, go one stage earlier than you think you are.
- Identify your primary metric. Each stage has one metric that matters most. Pick it. Put it on your homepage, your weekly standup, your investor updates.
- Set your milestone before you hit it. The milestones in Section 2 are thresholds, not celebrations. Define yours in advance so you're not moving goalposts retroactively.
- Run the anti-pattern check. Section 4 lists the most common metric mistakes by stage. Check yourself against it before your next update.
Section 1: The Five Stages and What Signal You're Actually Hunting
Every product moves through roughly five stages before it reaches scale. The boundaries are fuzzy. The metrics are not.
| Stage | Definition | Primary Question | What Signal You Need |
|---|---|---|---|
| 0. Pre-Build | You have a hypothesis, no product | Does this problem exist at the severity I think it does? | Qualitative density: Are 7 out of 10 conversations unprompted and emotional? |
| 1. Early Access / Alpha | Product exists, 5-25 users, mostly manual | Does anyone want this enough to come back? | Retention week 1: Did they use it again without you nudging them? |
| 2. Private Beta | Product works, 25-150 users, onboarding improving | Am I solving the right problem for the right person? | Activation rate + qualitative "aha" moment mapping |
| 3. PMF Search | Users paying or deeply engaged, iteration tight | Do I have a repeatable reason people stay? | Retention curve flattening (cohorts not going to zero) |
| 4. PMF Confirmed / Pre-Scale | Retention is solid, ICP is clear, economics are directional | Can I grow this without breaking it? | CAC payback period + net revenue retention |
| 5. Scale | Repeatable growth motion, team in place | Am I compounding or just spending? | LTV:CAC ratio, NRR > 100%, payback < 12 months |
The biggest strategic error founders make: staying in Stage 1 metrics (activation, engagement) while asking Stage 4 questions (can we grow?). You cannot answer a growth question with engagement data.
Section 2: The Metrics & Milestones Comparison Table
This is the full map. For each stage, here is what to measure, what milestone signals you've passed the stage, and what the leading indicator is that tells you before the lagging indicator confirms it.
| Stage | Track These Metrics | Milestone That Unlocks Next Stage | Leading Indicator to Watch | Vanity Trap to Ignore |
|---|---|---|---|---|
| Pre-Build | Interviews completed, problem-confirmation rate, willingness-to-pay signals | 15+ interviews with 70%+ unprompted problem description | People who follow up without being asked | Waitlist signups (they cost nothing) |
| Early Alpha | D1 / D7 retention, session count per user, support ticket themes | 40%+ of users return in week 2 without a nudge | Users who recruit other users | Total signups, total logins |
| Private Beta | Activation rate (hit the "aha" moment), qualitative satisfaction score, feature usage by cohort | 60%+ of new users reach the activation event within 72 hours | NPS from week-2 cohorts (not all-time) | Average NPS across all users ever |
| PMF Search | Weekly/monthly retention by cohort, engagement depth (not breadth), churn reasons by segment | Retention curve flattens above 30% at month 3 for a defined ICP | Second-order retention (do churned users come back?) | Total MRR if it's growing from bad-fit customers |
| PMF Confirmed | CAC by channel, time-to-value, net revenue retention | NRR > 100% for 2+ consecutive quarters | Expansion MRR as a % of new MRR | Gross revenue without retention context |
| Pre-Scale | CAC payback period, sales cycle length, ICP win rate, onboarding completion rate | Payback period < 18 months, win rate stable across sales reps | Pipeline velocity (not just pipeline size) | Number of enterprise logos without contract value |
| Scale | LTV:CAC by channel, payback period trend, NRR, gross margin | LTV:CAC > 3x, payback trending down, NRR > 110% | CAC efficiency ratio improving quarter over quarter | Total users, total revenue without margin context |
One note on this table. These thresholds are directional, not universal. SaaS, marketplace, and consumer products have different benchmarks. What does not change is the logic: retention before growth, activation before retention, problem confirmation before activation.
Section 3: Templates
📋 Weekly Metrics Update Template (for internal use and investor updates)
## Week [N] Update — [Company Name]
**Stage**: [Pre-Build / Alpha / Beta / PMF Search / PMF / Scale]
**Primary metric this stage**: [One metric only]
**This week's number**: [Value]
**Last week**: [Value]
**Direction**: [Up / Down / Flat — and why in one sentence]
---
**Activation**
- New users this week: [N]
- Activation rate (reached core event): [X%]
- Time-to-activation median: [X hours/days]
**Retention**
- D7 retention (most recent cohort): [X%]
- D30 retention (30-day-old cohort): [X%]
- Notable change in cohort behavior: [one sentence]
**Revenue (if applicable)**
- New MRR: $[X]
- Expansion MRR: $[X]
- Churned MRR: $[X]
- Net new MRR: $[X]
- NRR (trailing 90 days): [X%]
---
**What we learned this week**: [2-3 sentences max. What changed your thinking?]
**One thing we're changing because of it**: [One sentence]
**What we're not changing**: [One sentence — shows conviction]
---
**Milestone status**
- Target milestone: [Describe it]
- Current progress: [X% / on track / at risk]
- Expected date: [Month]
📋 Investor Metrics Brief Template (board meetings and fundraise prep)
## [Company] — Metrics Brief [Quarter / Month]
**Stage**: [Self-assessed, be honest]
**Fundraise status**: [Not raising / Raising / Closed]
---
### The one number that matters right now
[State your primary metric, its current value, its value 90 days ago, and the trend line in 2 sentences.]
---
### Retention
| Cohort | D7 | D30 | D90 |
|--------|-----|-----|-----|
| [Month 1] | X% | X% | X% |
| [Month 2] | X% | X% | X% |
| [Month 3] | X% | X% | X% |
**What the curve tells us**: [One sentence interpretation]
---
### Revenue (if applicable)
- MRR: $[X] (was $[X] 90 days ago)
- NRR: [X%]
- CAC (by primary channel): $[X]
- CAC payback: [X months]
- LTV estimate: $[X] (basis: [explain assumption])
---
### Unit economics narrative
[2-3 sentences. Not just numbers. What do the numbers mean about the business? Where are they going?]
---
### What we're not tracking yet (and why)
[Investors respect founders who know what they don't know. Name 1-2 metrics you'll add next stage and when.]
---
### Milestone we're working toward
[State it plainly. What is the specific number or event that marks the end of this stage?]
**Confidence**: [High / Medium / Low] — [One sentence on basis of confidence]
📋 PMF Diagnostic Template (run this every 90 days)
## PMF Diagnostic — [Date]
### Step 1: Retention test
- Pull your month-3 retention for the cohort from 3 months ago.
- Is it above 25% (B2B SaaS) or 20% (consumer)? [Y/N]
- Is the curve flat (not still declining)? [Y/N]
If both are yes: you have a retention signal. You don't yet have PMF confirmed.
---
### Step 2: Qualitative test
- In your last 10 user interviews, how many described the problem without being prompted? [N/10]
- How many said they'd be "very disappointed" if the product went away? [N/10]
- How many have referred at least one other user? [N/10]
If 7+ on any of these: strong qualitative signal.
---
### Step 3: Expansion test (B2B)
- Of your paying customers from 6+ months ago, what % have expanded? [X%]
- Is your expansion MRR positive (more upgrades than downgrades)? [Y/N]
---
### Step 4: Honest ICP statement
Write one sentence: "We have strongest retention with [persona] who use us for [use case] and experience [outcome]."
If you can't write this sentence with confidence, you don't have PMF. That's fine. That's the work.
---
### Summary verdict
- Retention signal: [Yes / Partial / No]
- Qualitative signal: [Yes / Partial / No]
- Expansion signal: [Yes / Partial / No / Too early]
- PMF status: [Confirmed / In search / Too early to know]
Section 4: System Prompts for Metric Analysis
🤖 Prompt: Diagnose your current stage
You are a product strategy advisor helping a founder understand what stage their product is in and what metrics they should be prioritizing.
I'll describe my current situation and you will:
1. Identify which stage I'm in (Pre-Build, Alpha, Beta, PMF Search, PMF Confirmed, or Scale)
2. Name the one metric I should treat as primary right now
3. Identify any metrics I'm currently tracking that are likely misleading me for this stage
4. Give me the milestone I should be working toward before thinking about the next stage
Be direct. If I'm in a stage earlier than I think, say so. Founders regularly misidentify their stage because they're optimistic.
My situation: [describe your product, how long you've been live, how many users/customers, what you're currently tracking, and what feels broken]
🤖 Prompt: Interpret a retention curve
You are a product analyst helping me understand what my retention data is telling me.
I'll give you cohort retention data and you will:
1. Describe what the curve shape means in plain terms
2. Identify whether the curve has flattened (and at what level)
3. Tell me what this means for my PMF status
4. Suggest 2-3 hypotheses for why retention looks the way it does
5. Name the one thing I should investigate or change first
Do not give me generic retention advice. Interpret the data I give you specifically.
My data: [paste cohort table or describe it — e.g., "month 1 cohort: D7 = 45%, D30 = 28%, D60 = 22%, D90 = 21%"]
My product: [one sentence description]
My ICP: [describe who these users are]
🤖 Prompt: Write a metrics narrative for investors
You are a fundraising advisor helping a founder communicate their metrics clearly and credibly.
I'll give you my raw numbers and you will write a metrics narrative that:
1. Opens with the most important number and its trend
2. Explains what the unit economics imply about the business
3. Names what the data does not yet show (and why that's appropriate for my stage)
4. Ends with the milestone I'm working toward and why it matters
Do not hype or spin. Investors read hundreds of decks. Honest framing with clear logic is more compelling than optimistic framing with weak logic.
My numbers: [paste your metrics]
My stage: [self-assessed]
My raise: [amount / target close date if applicable]
Section 5: Common Pitfalls
Tracking MRR before you have retention. Revenue from customers you can't retain is a loan, not a business. If your month-3 retention is heading to zero, your MRR number is telling you a lie.
Using all-time aggregates instead of cohorts. An aggregate NPS or retention number hides everything that matters. Cohorts show you whether you're getting better. Aggregates show you what happened on average across every mistake you ever made.
Calling a feature metric a business metric. "Users who use Feature X retain at 80%" is not a business insight until you know what % of users reach Feature X and whether that path is reproducible.
Setting milestones after you hit them. Moving the goalpost forward is fine. Redrawing it retroactively so a number you hit counts as a milestone you planned is how you confuse yourself into thinking you're further along than you are.
Optimizing for investor-friendly metrics before you have product-market fit. If you're padding MRR with annual contracts from warm intros before you've solved retention, you're borrowing time. Investors who know what they're doing will find the retention data.
Reporting growth without reporting the denominator. "We tripled our user base" from 10 to 30 users is a different sentence than the same statement from 1,000 to 3,000. Always give the absolute number.
Treating NPS as a health metric at the wrong stage. NPS is a useful signal after you have enough users to make it statistically meaningful and enough time to act on it. Asking 12 beta users for NPS and averaging it is not a business metric. It's a conversation prompt.
Why We Built This
Coding is becoming cheaper. Knowing what to build is becoming more valuable. But knowing what to build is not just a strategy problem. It's a measurement problem. If you can't identify the right signal at the right stage, you make decisions based on noise.
ProductOS is built around this. Most tools start at "how to build." We start at "what to build" and why. That means research, problem definition, and prioritization happen before a line of code is written, and the context from those decisions carries through all the way to deployed product. Metrics and milestones are not an afterthought in that process. They're part of the definition of what "done" looks like.
This guide is a standalone resource. Use it without ever touching our product. If it helps you get clearer on what you're measuring and why, it's done its job.
If any of this lands and you want to see it in action, we're at productos.dev. No pressure. The toolkit stands on its own.
If you'd rather have humans plus AI run this for you on a real product today, that's what 1Labs AI does.
Build it with ProductOS
Stop reading about it. Ship it.
Describe your idea once. AI agents research it, spec it, design it, and build real code you own, sharing one context the whole way.
Free to start · no credit card required
Founder & CEO, ProductOS
CS engineer and IIM Lucknow MBA. Built products across enterprise and AI for 10+ years. Founded ProductOS to give every PM and founder the leverage of a full product team. Writes about AI product development, PRDs, and building with agents.