Instagram creator Liz (@liz.on.the.web) handed her entire account strategy to Claude for one week. The results were staggering: 1.4 million views and $17,444 in revenue, all without filming a single video or showing her face. She used just four carefully crafted prompts to research, script, write, and automate her content pipeline.
This breakdown walks through each of those four prompts and explains why they work so well for AI powered Instagram growth.
What Claude Prompts Actually Drive Instagram Viral Content?
The secret behind Liz's results was not a single magic prompt. It was a four step system where each prompt built on the output of the previous one. The workflow moved from research to scripting to hook writing to automation, creating a repeatable content engine.
Understanding how the Instagram algorithm rewards engagement patterns is key to why this approach works. Claude analyzed what was already performing well, then reverse engineered the winning formulas.
Step 1: Find What Actually Goes Viral in Your Niche
The first prompt focused entirely on research before creation. Instead of guessing what might work, Liz used Claude to analyze existing high performing content across multiple platforms.
Here is the prompt structure:
Analyze the highest performing Instagram Reels, TikToks, and Reddit posts in the [niche] niche from the last 30 days. Identify repeating hooks, visual styles, emotional triggers, and content formats that consistently generate high engagement. Then summarize the 5 strongest content angles optimized for AI generated content and short form videos.
This prompt works because it forces Claude to ground its output in real data rather than generic advice. By specifying a 30 day window, you get current trends instead of outdated tactics. The multi platform analysis reveals cross platform patterns that most creators miss when they only look at Instagram.
Pair this research step with a strong Instagram content strategy to turn insights into a posting calendar.
Step 2: Build Scripts That Keep People Watching Until the End
The second prompt took the winning content angles from step one and turned them into full scripts. The goal was maximizing watch time, which is the single most important metric the algorithm uses to distribute Reels.
The script building prompt asked Claude to structure content with a strong open loop in the first two seconds, escalating tension through the middle, and a satisfying payoff at the end. This mirrors the narrative structure that top creators use instinctively.
Key elements the prompt optimized for:
- Hook within 1.5 seconds: The opening line must create a curiosity gap
- Pattern interrupts every 5 to 8 seconds: Prevents the swipe away
- Emotional escalation: Each sentence raises the stakes slightly
- Clear payoff: The viewer needs to feel rewarded for staying
This approach directly impacts your engagement rate because longer watch times signal quality to the algorithm.
How Do You Write Hooks That Stop the Scroll?
The third prompt was dedicated entirely to crafting scroll stopping hooks. Liz's carousel emphasized that Prompt 3 is where most creators fall short. A great script with a weak hook still gets ignored.
The hook writing prompt generated multiple variations for each piece of content, testing different psychological triggers:
- Curiosity gaps: "Nobody talks about this Instagram feature..."
- Contrarian statements: "Everything you know about viral content is wrong"
- Specific numbers: "This one change added 40K followers"
- Pattern interrupts: Starting with an unexpected word or visual
The carousel itself used these exact principles. Slide one opened with "I gave my Instagram to Claude" followed by the specific results "1.4M views, $17,444 in one week." The combination of a bold claim with precise numbers creates an irresistible curiosity gap.
Step 4: Put the Whole Thing on Autopilot
The fourth prompt was the one Liz described as "the one nobody shows you." It focused on automating the entire content pipeline so the system could run without daily manual input.
This prompt configured Claude to handle the full cycle: researching trends, generating scripts, writing hooks, scheduling posts, and even crafting reply templates for comments. The automation extended to DM automation for handling the flood of engagement that viral content generates.
The automation setup meant that once the system was running, Liz could step back while the prompts continued producing content on schedule. This is the difference between a creator who works in their business and one who builds a system that works for them.
Why Does AI Generated Content Outperform Manual Creation?
AI generated content outperforms manual creation for several reasons that Liz's results demonstrate clearly:
Speed of iteration: Claude can generate 20 script variations in the time it takes a human to write one. This means more testing, faster learning, and quicker identification of what resonates.
Data driven decisions: Instead of relying on gut feeling, the prompts forced every creative decision to be backed by analysis of what was already working in the niche.
Consistency: Human creators have good days and bad days. A well configured prompt system produces consistent quality output regardless of mood, energy, or inspiration.
Scale: One person using these four prompts can produce the content volume of a small team. Liz generated enough content in one week to accumulate 1.4M views without filming anything.
Using a tool like InstantDM's growth service alongside these prompts amplifies the results further by handling audience building while the content engine runs.
What Makes This Prompt System Different and How It Drives Revenue
Most people who use AI for Instagram content ask generic questions like "write me an Instagram caption about marketing." This produces generic results that get generic engagement.
Liz's four prompt system works differently because each prompt has a specific job in the pipeline. The research prompt feeds the script prompt. The script prompt feeds the hook prompt. The automation prompt ties everything together. It is a system, not a collection of disconnected requests.
The carousel slides reinforced this point by showing the actual prompt text with specific instructions about niche targeting, time frames, and output formats. Vague prompts produce vague results. Specific prompts produce specific, actionable content.
That specificity translated directly into revenue. Liz's $17,444 in one week came from the compounding effect of viral reach. When content gets 1.4M views, even modest monetization strategies produce significant revenue through affiliate commissions, digital product sales driven by the comment to DM funnel, and sponsored post opportunities. The research prompt in step one ensured every piece of content targeted an audience with buying intent.
How to Set Up Your Own Claude Content System
To replicate this approach, start with these principles:
- Define your niche precisely: The more specific, the better Claude can research
- Feed real data: Ask Claude to analyze actual top performing posts, not hypotheticals
- Chain your prompts: Each output should feed the next prompt as input
- Test hook variations: Never settle for the first hook Claude generates
- Automate the repeatable parts: Once the system works, remove yourself from the daily execution
The creators who win with AI content are not the ones with the best single prompt. They are the ones who build a system where prompts work together to produce consistent, high performing content at scale.
Start by testing these four prompts in your own niche. Track what the algorithm rewards using an engagement rate calculator and refine your prompts based on real performance data. The results may surprise you.
