---
title: "Build a repeatable growth loop"
description: "Move from first users to measured, sustainable acquisition."
---
Growth is a loop: reach the right people, help them experience value, retain them, and use what you learn to improve the next acquisition cycle. Scale only after the loop works at a small volume.

## Instrument the complete journey

```
impression → App Store view → download → onboarding → activation
→ trial or purchase → retained use → referral or renewal
```

Give every stage one primary metric. Diagnose from the top: if qualified people never reach the product page, onboarding changes cannot solve the immediate problem.

## Grow in stages

### 1. Prove the promise

Interview and recruit users directly. Confirm that they reach and repeat the core result.

### 2. Find one acquisition signal

Publish original content, work with a small creator, or join a relevant community. Look for one message and audience combination that produces activated users.

### 3. Make the result repeatable

Reproduce the winning concept several times. Document its audience, hook, format, conversion path, and cost before increasing volume.

### 4. Scale with guardrails

Increase spend or output gradually while monitoring acquisition cost, trial conversion, retention, refunds, support volume, and proceeds after platform fees.

## Know the unit economics

Compare the cost to acquire a paying customer with the contribution that customer produces after refunds, fees, creator payments, and variable service costs. Use conservative retention assumptions; early revenue alone can hide an unprofitable channel.

## Run a weekly growth review

-   Which audience and message produced the most activated users?
-   Where did the largest qualified-user drop occur?
-   Which cohort retained or converted better after seven and thirty days?
-   What single experiment will reduce the largest uncertainty next week?

:::tip
A revenue target is an outcome, not a plan. Build the plan from acquisition volume, conversion, retention, price, and cost assumptions you can measure separately.
:::