The 7 Metrics That Tell You If Your MVP Is Working

Introduction: The Vanity Metric Trap

Congrats - your MVP is live. You’re staring at dashboards:

1,000 pageviews. 200 sign‑ups. 50 downloads.

Feels good, right? The problem: none of these prove your MVP is working.

What matters isn’t traffic or downloads. What matters is whether users complete the flow you designed - and whether money or value is exchanged. This post gives you the 7 MVP metrics that actually matter, how to instrument them in a day, and how to read them without fooling yourself.

The guiding idea: measure your thin slice. If your MVP is “Upload transcript → Generate report → Share,” then your metrics live along that path. Not on your homepage.

---

1) First Contact

Definition: the very first measurable touchpoint (e.g., landing page visit with a UTM, ad click, referral link open, App Store view).

Why it matters: shows which channels create awareness. You can’t improve conversion if you don’t know the entry points.

Examples:

• Landing page pageview with utm_source and utm_campaign
• App Store listing view (impressions)
• Link click from a partner email using a tracked parameter

How to instrument:

• Add a single event: first_contact with properties: source, campaign, medium, creative, and an experiment flag if A/B testing.
• Capture timestamp and a simple anonymous user id (cookie or device id) before sign‑up.

What good looks like (week 1):

• 3–5 channels sending any traffic
• One clear winner on click‑through rate or cost per first contact

Pitfalls:

• Counting bots or internal views (filter your IPs and preview domains)
• Overfitting to impressions without downstream conversion

---

2) Activation

Definition: the moment a user starts the core flow - the first action that meaningfully commits them. Think “created first project,” “uploaded first file,” “connected Stripe,” “added first item.”

Why it matters: separates the curious from the committed. A strong activation rate often predicts everything that follows.

Examples:

• Project tool: project_created
• Media tool: file_uploaded
• Finance tool: connected_bank or connected_stripe

How to instrument:

• Choose a single activation event and put it behind sign‑up if possible (so it’s tied to a user id).
• Capture minimal properties that shape onboarding (plan, role, import method).

Benchmarks (directional):

• 25–60% of sign‑ups should activate on Day 1 for a simple MVP with a clear task.

Pitfalls:

• Defining activation as “clicked around.” Pick a concrete, irreversible action.
• Hiding activation behind too many steps; reduce friction to reach it.

---

3) Completion

Definition: the user finishes the core flow you promised. If activation is “started,” completion is “done.”

Why it matters: proves your MVP delivers value. This is the heartbeat metric.

Examples:

• Invoicing: invoice_generated
• Support: ticket_resolved
• Commerce: checkout_complete
• Analytics: report_ready

How to instrument:

• Emit a single completed_core_flow event (or your domain‑specific event) with a reference to the activation object (e.g., project_id, file_id).
• Add duration: time from activation to completion in seconds. If you can’t compute server‑side, capture client timestamps at both steps and subtract.

Completion rate:

• Define a simple funnel: Activation → Completion
• Target an initial 30–70% completion for narrow MVPs; if you’re <20%, you have friction or unclear value.

Pitfalls:

• Counting partial success as completion.
• Ignoring errors/timeouts that block completion (log a core_flow_error with error class and surface area).

---

4) Conversion (Money Moment)

Definition: someone pays - subscription, one‑time purchase, credit top‑up, or pilot invoice.

Why it matters: strongest validation. Even $1 is a stronger signal than 1,000 free users.

Examples:

• payment_succeeded (Stripe event mirrored into your analytics)
• plan_upgraded
• invoice_paid

How to instrument:

• Use your payment provider’s webhooks as the source of truth; forward a sanitized event into product analytics.
• Store amount, currency, plan, and whether it’s test or live mode.

Conversion rate:

• Don’t obsess over “Sign‑up → Pay” on Day 1 unless your product is priced for impulse buys. For B2B MVPs, track “Completion → Pay” within 14–30 days.

Pitfalls:

• Counting trial starts as conversion; separate trial_started from payment_succeeded.
• Double‑counting renewals as new conversions.

---

5) Retention

Definition: do they come back a second time? Easiest form: repeat activation or repeat completion within a time window (D7/D30).

Why it matters: sticky MVPs get used again. If users vanish after one try, you may have curiosity, not demand.

Ways to define quickly:

• Same user triggers activation or completion again within 7 days (D7) and 30 days (D30).
• Or: user performs any “active use” event on 2+ distinct days in the first 7 days.

Targets (directional, MVP stage):

• D7 Retention: 15–35% for tools that fit weekly workflows; higher for daily utilities.
• D30 Retention: 10–25% for early B2B tools; consumer can vary widely.

Pitfalls:

• Using logins as retention; measure meaningful actions.
• Comparing across products with different cadences (weekly vs monthly jobs).

---

6) Referral / Word of Mouth

Definition: do users bring in others, or does usage create invites inherently?

Why it matters: when people tell friends, the problem is real - and your product likely solved it well enough to mention.

Quick measures:

• % of sign‑ups with a non‑paid referral source (utm_source=referral, ?ref= params, invite links)
• Count of invite_sent and invite_accepted
• K‑factor proxy: invites accepted per active user in a week

How to instrument:

• Issue per‑user invite links; capture referrer_user_id on sign‑up.
• Add a lightweight “share” loop (copy link, email invite) in the completion screen.

Pitfalls:

• Mistaking partner traffic for organic word of mouth; tag partner sources separately.
• Incentivizing low‑quality referrals (e.g., giveaways that don’t correlate to use).

---

7) Speed of Feedback

Definition: how quickly do you learn? Measure the loop from event → insight → change in the product.

Why it matters: MVPs die when feedback loops are too slow. If it takes six months to know if something worked, that’s not an MVP.

Three simple cycle‑time metrics:

• Time to first value (TTFV): sign‑up → completion; aim to shrink this every week.
• Time to fix: bug opened on core flow → fix shipped.
• Time to decision: insight logged → change shipped behind a flag.

How to instrument:

• Track timestamps on completion and on deploys to compute TTFV trend.
• Maintain a lightweight “insight log” (Notion/Sheet) with created_at and shipped_at; compute median.

Pitfalls:

• Over‑engineering; you don’t need Jira dashboards. A sheet works for Sprint 1.

---

Cross‑Cutting Signals That Help You Decide Faster

These aren’t part of the “7,” but they sharpen decisions:

• Time to set up: account creation → first activation; if this is >10 minutes for a simple tool, you have friction.
• Error budget: weekly count of core‑flow errors; trend down aggressively.
• Support p95 response time: if you offer pilots, reply fast; product quality feels higher when support is tight.

---

Instrumentation Playbook (Day‑1 Setup)

You can wire basic product analytics in under an hour. Keep it boring.

• Pick one tool: PostHog, Amplitude, or a minimal custom events table.
• Define 7 events: first_contact, activated, completed_core_flow, payment_succeeded, invite_sent, invite_accepted, core_flow_error.
• Attach minimal properties only: source/campaign, plan, role, object ids, amounts, and durations.
• Standardize user ids immediately after sign‑up; backfill anonymous id as an alias.

Event naming tips:

• Use past‑tense verbs (created, uploaded, generated).
• Keep names domain‑specific where helpful (e.g., report_generated) but map them to the 7 concepts in your own cheat sheet.

Server vs client:

• Client: first contact, activation click, UI errors.
• Server: completion (safer), payments, durable failures.

---

Reading the Numbers: A Simple Weekly Ritual

Every Friday, review a one‑page doc or dashboard with:

• Volumes by step: First contact → Activation → Completion → Conversion
• Rates: Activation rate, Completion rate, Pay rate
• Retention: D7 and D30 for cohorts that reached completion
• Referrals: % of new sign‑ups from invites/ref links
• Speed: median TTFV; median time to fix core‑flow bugs

Ask three questions:

1. Where is the biggest drop‑off relative to our promise?
2. What is the cheapest change we can make this week to move one number?
3. What did we learn from support tickets, emails, and calls that aligns with the numbers?

Ship one change per week to move one metric. Small levers compound.

---

Case Study: Transcript → Report MVP (30 Days)

Context: a lightweight tool that turns audio transcripts into a structured report.

Plan: 14‑day sprint. Promise: “Upload a transcript and get a publish‑able report in under 10 minutes.”

Metrics after 30 days:

• 30 sign‑ups (first contact) - from a mix of LinkedIn posts and a small newsletter mention
• 12 uploads (activation)
• 10 reports generated (completion)
• 2 Stripe payments (conversion)
• 7 second sessions (retention)
• 3 invites sent; 1 accepted (referral)
• Median time to first value: 8m 40s (speed)

Reading it:

• Activation is low (12/30) relative to sign‑ups; onboarding likely unclear.
• Completion is good (10/12) once started - the core value lands.
• Conversion is small but real (2 paid) - prioritize tightening activation and checkout friction.

What we changed in Week 5:

• Added a 60‑second sample video + demo transcript to eliminate “blank page” fear.
• Moved pricing clarity into the flow (show price before upload, not after).
• Added a one‑click invite at the success screen with a pre‑filled email.

Week 6 results:

• 25 sign‑ups → 18 uploads → 15 completions → 5 payments; D7 retention moved from 23% to 31%.

Lesson: work the bottleneck you can reach fastest. Here, activation.

---

How to Build Your MVP Metrics Dashboard (Template)

You don’t need Looker to start. A Notion or Google Sheet is enough. Here’s a simple structure you can copy.

Sheet tabs:

• Inputs: paste weekly counts from your analytics tool (or export CSVs)
• Cohorts: a simple table of users by week of completion and whether they came back by D7/D30
• Funnel: a weekly funnel with counts and rates
• Notes: your Friday decisions and shipped changes

Columns to include:

• week_start (date)
• first_contact (count)
• activated (count)
• completed (count)
• paid_users (count)
• d7_retained (count)
• d30_retained (count)
• referrals (count)
• invites_accepted (count)
• ttfv_seconds (median)

Computed fields:

• activation_rate = activated / signups
• completion_rate = completed / activated
• pay_rate = paid_users / completed (or signups, depending on model)
• d7_retention_rate = d7_retained / completed
• k_factor_proxy = invites_accepted / activated
• ttfv_minutes = ttfv_seconds / 60

Weekly review checklist:

• Validate data sanity (no sudden zeros due to tracking)
• Identify the steepest funnel drop
• Pick one change to ship next week; write it in Notes with an owner
• Set a micro‑target (e.g., +10% activation) and review next Friday

---

Common Traps (And How to Avoid Them)

• Chasing “north star” buzzwords before you have a working core flow. Your north star is completion, not MAUs.
• Counting pageviews and logins as success. Only meaningful actions matter.
• Measuring too much. Seven numbers are plenty in Sprint 1.
• Delaying instrumentation. Add events now; you can refactor names later.
• Ignoring qualitative signals. Numbers tell you where; conversations tell you why.

---

Category‑Specific Notes (Quick Hits)

• Marketplaces: Activation may be two‑sided. Track seller activation and buyer activation separately; define completion on both sides (listing created; purchase completed). Measure time to first transaction.
• Productivity tools: Retention is king. Tie retention to a recurring job (weekly report, daily standup). Add heartbeat nudges only after the core flow works.
• Fintech: Conversion must reconcile with provider webhooks; treat payment provider as the source of truth. Add a reconciliation_mismatch alert if your counts drift.
• AI‑assisted tools: Measure “assisted edit” rate - how often the user accepted or lightly edited the AI output, not just generated content.

---

Implementation Notes for Non‑Technical Founders

• Ask your builder to wire the seven events as part of the MVP definition of done.
• Request a 1‑pager with event names, properties, and where they’re emitted.
• Insist on a Day‑14 demo with a mini dashboard (can be a Sheet) showing the seven numbers.

If a vendor can’t deliver basic instrumentation, they’re not shipping a production‑ready MVP.

---

Key Takeaway

The right metrics prove demand. Forget vanity downloads. Focus on completion, conversion, and retention - and how fast you learn. If you track these seven from Day 1, you’ll know exactly where to push next week.

---

The MVP Metrics Dashboard (Lead Magnet)

I built a Notion/Sheet template with these 7 metrics:

• First contact
• Activation
• Completion
• Conversion
• Retention
• Referral
• Feedback loop

👉 Download the MVP Metrics Dashboard Template here.

---

CTA

Grab the MVP Metrics Dashboard Template and book a 20‑min scope call.

Visit weekonelabs.com.