Grow On Instagram

---
name: grow-on-instagram
description: "Build a rigorous Instagram growth system for public-facing creators and brands: choose a niche and differentiation, develop Reel topics, hooks, formats, visual layouts, storytelling, series, trust, and conversion paths, then improve them from performance data. Use for Instagram or Reels strategy, content calendars, scripts, hook rewrites, format selection, retention audits, creator-brand systems, or post-performance analysis."
---

# Grow on Instagram

Design a learning system that earns repeat attention from a defined audience. Do not promise reach, followers, or revenue; use observed results to form and test hypotheses.

## Start with the operating brief

Define the audience, their desired change, the narrow topic space, the conversion goal, available proof, production constraints, and existing metrics. Use one niche/topic space at the outset; express personality through the treatment, perspective, and proof rather than mixing unrelated subjects.

Choose one primary differentiation to develop first:

- **Visual experience:** more distinct, readable, or demonstrable than competing posts.
- **Ideas and angles:** more surprising, useful, or clearly reasoned.
- **Value and proof:** more actionable substance, demonstrations, or credible results.

Read [growth-and-measurement.md](references/growth-and-measurement.md) before planning a new account or a multi-week system.

## Build the content matrix

1. Select 3–5 audience-relevant topic pillars backed by a genuine point of view or proof.
2. Assign 2–3 formats to each pillar. Start with formats the creator can reliably execute; add one more ambitious format as a deliberate test.
3. Assign 2 visual-layout treatments to each topic/format combination.
4. Give the output a recurring series promise. A series should help a viewer understand what to expect and what to watch next.
5. Schedule a test cycle. Change one meaningful variable at a time—topic, hook, format, proof, visual layout, or CTA—so the outcome remains interpretable.

Use [hooks-formats-and-angles.md](references/hooks-formats-and-angles.md) for the actual creative choices.

## Write every Reel around one tension

Complete this brief before scripting:

```markdown
Viewer: [specific person]
Topic / pillar: [one subject]
Tension: [wrong belief, unresolved question, friction, or desired outcome]
Single promise: [what becomes clearer or easier]
Angle: [new reveal, contrast, insider view, proof, connection, or future change]
Evidence: [demo, example, result, story, observation, or cited source]
Hook stack: [on-screen text] + [spoken line] + [first visual]
Open loop: [question the viewer wants resolved]
Value beats: [2–4 escalating beats]
Re-hook: [new question, contrast, reveal, or visual reset]
Payoff: [specific answer, process, or decision]
CTA: [one contextually useful next step]
Series / next view: [what to watch or do next]
```

Make the on-screen text legible without sound. Align it with the spoken line and first visual; do not make the viewer reconcile competing messages.

Read [storytelling-and-retention.md](references/storytelling-and-retention.md) for open loops, novelty, re-hooks, stakes, and proof.

## Design for memory and depth

Assume a feed viewer may see only one post. Give the post a complete payoff, then create a voluntary path to depth: a related series, profile, email resource, long-form piece, product, or community. Use recognizable objects, aesthetics, supporting characters, and cross-surface continuity only when they authentically fit the brand.

Build trust through the strongest evidence the creator can honestly provide. Show the work, explain tradeoffs, and make claims proportionate to the proof. Never manufacture testimonials, urgency, authority, scarcity, or endorsements.

## Run the weekly learning loop

Record each post’s topic, angle, format, hook stack, visual treatment, proof type, story mechanism, CTA, and post age. Compare like with like. Track views, retention/watch time where available, shares, saves, comments, profile actions, follows, leads, and conversion quality.

Classify results:

- **Keep:** retention and a meaningful downstream signal support the same mechanism.
- **Iterate:** one element worked; retain it and test one adjacent variable.
- **Stop:** repeated comparable tests underperform without another strategic reason to continue.

Maintain a swipe file of mechanisms, not scripts. When reviewing competitors, extract the audience tension, hook type, proof, structure, and visual treatment; then create an original expression for the current account.