<aside> 🧠

What you get: an input check, a confirmed funnel, 15–20 ranked opportunities, and 10 test briefs. Everything runs on your real connected data. It never invents a number and never touches your store (read-only).

</aside>

<aside> 🚀

Want this run on your store instead of running it yourself? We do this exact audit for DTC Shopify brands doing $50M to $300M, then ship the tests. +10% Revenue Per Visitor in 90 days. Or we refund the engagement. Book a teardown →

</aside>

What's inside

• The exact connector and export setup for Shopify, GA4, Clarity, PageSpeed, and experiment history
• A preflight that proves every source belongs to the right store before analysis starts
• One copy-and-paste master prompt plus four deeper audit prompts
• A visual route for live storefronts, heatmaps, and page screenshots
• A 30-minute first pass once the inputs are ready
• A clean format for 15–20 opportunities and 10 test briefs

Copy this first: the master prompt

Analyze my heatmap, GA4, and live store.

Run this as a read-only CRO audit.

Before analysis:
1. Identify the connected Shopify shop, GA4 property, and Clarity project.
2. Show the exact date range and source beside every metric.
3. Stop any source that cannot be matched to this store. Never substitute another property.
4. Use structured data for exact values and images for visual context.
5. Label conclusions OBSERVED, INFERRED, ASSUMPTION, or MISSING.
6. Do not invent conversion rates, revenue, lift, or test duration.
7. Do not change the store or any connected app.

Then:
- scan the live storefront on mobile and desktop
- build the approved GA4 purchase funnel
- inspect the matching Clarity heatmap and tap export
- cross-check prior experiment results
- rank 15–20 distinct opportunities
- write the top 10 as test briefs with Evidence, Hypothesis, Primary metric, Guardrail, Dependencies, and QA notes

Start with the input check. If a live source fails, state the blocker once and use its approved export fallback.

<aside> ✓

Do not skip the input check. A connector being installed does not prove it is authenticated, pointed at the right store, or returning the fields you need.

</aside>

The workflow

flowchart LR
    A["Shopify + GA4<br>Clarity + PageSpeed<br>store and competitor URLs"] --> B["ChatGPT checks<br>the inputs"]
    B --> C["Finds the biggest<br>funnel leaks"]
    C --> D["Explains likely<br>page friction"]
    D --> E["Ranks 15 to 20<br>opportunities"]
    E --> F["Writes 10<br>test briefs"]

Two things matter:

• ChatGPT can speed up the first pass. It still needs clean data.
• The output is a roadmap, not proof. An opportunity becomes a win only after a valid test.

<aside> ✓

The rule: ChatGPT does the sorting, comparison, and first draft. You keep the decisions.

</aside>