coreyhaines31/marketingskills / skills/reverse-engineer-marketing
reverse-engineer-marketing
VerifiedWhen the user wants to know how another company grew and what to copy, or asks what a competitor did and when: a dated log of their moves against their traffic from public data, archives, job posts, reviews, and their funnel, compared with the user's data, ending in plays routed to other skills. Also finds where competitors are weak. For a snapshot profile, see competitor-profiling. For comparison pages, see competitors.
references/funnel-hack.md
View on GitHub# Funnel Hack
Funnel Hack mode in full: buy from the target the way a customer would, and map every step from the first ad to the last follow-up. It extends [go-inside.md](go-inside.md), and all of its rules apply: the user approves each action, the user's real identity, and a dedicated inbox on the user's domain.
## Contents
- Before you buy
- Map the money path
- Find their tests
- Read their paid search
- Build the swipe file
- The funnel map deliverable
---
## Before You Buy
- **Set a budget with the user** before the first purchase: the total, the products, and whether to take upsells. Every purchase needs an explicit yes with the amount.
- **Buy to keep.** Refund only under the seller's published policy and for a real reason. Don't plan purchases around refunds.
- **Use the user's payment method and name.** No virtual cards in made-up names.
- **Don't exploit bugs.** If a checkout misprices something or a discount stacks by mistake, note it and don't use it.
- **Pick the entry point a real buyer would use:** an active ad (from `sc-ads` or the ad libraries), a lead magnet, or the homepage. Record which one, because funnels differ by entry.
## Map the Money Path
Screenshot and record each step in `teardowns/<target>/funnel/steps.csv`:
`step,url,screenshot,offer,price,anchor_price,cta,timer_or_scarcity,notes`
| Step | What to capture |
|---|---|
| **Ad** | Hook, offer, and the exact landing page with its UTM parameters |
| **Landing page** | Headline, offer stack, proof, guarantee, CTA, page speed |
| **Opt-in** | Fields, lead magnet, consent language, SMS opt-in |
| **Thank-you page** | The immediate next offer, often a low-price first purchase |
| **Checkout** | Order bump, payment options (Shop Pay, buy now pay later, PayPal), shipping threshold, subscribe-and-save, trust badges |
| **Upsells and downsells** | Each one-time offer after checkout: product, price, and what happens if you decline |
| **Confirmation** | Post-purchase offers, referral prompt, account creation |
| **Follow-up** | Every email and SMS for 30 days, logged as in go-inside.md |
Then run the side paths:
- **Abandoned cart:** add to cart, enter the email, leave. Log the emails and texts, their timing, and whether the discount grows with each one.
- **Abandoned opt-in:** start the quiz or opt-in and stop halfway, if the flow captures the email early.
- **Popups:** note what triggers each one (time, scroll, exit intent), the offer, and whether it changes on a second visit (hand to **popups**).
- **Webinar funnels:** registration page, confirmation, reminder sequence, the replay page, the minute the pitch starts, the offer, and the deadline.
- **Quiz funnels:** every question, the branches, each result page, and the offer each result gets.
### Check the Urgency
Open each countdown timer, "only N left", and "offer ends tonight" message in two clean browser sessions, a few hours apart, and again after the stated deadline. Record exactly what you saw as Observed: "the timer restarted in a new session", "the count stayed at 7 for three days". Whether the scarcity is staged is Inferred: a count that never moves can mean no sales, and a restarting timer can be a disclosed per-visitor offer. Either way, don't copy urgency as if it's what converts, and remember that false scarcity runs into the FTC's rules on deceptive practices.
## Find Their Tests
- **`teardown.mjs experiments <domain>`** lists the client-side testing tools and, for Optimizely and GrowthBook, the readable names in their configs. Names like `pricing_annual_default_v3` or `checkout_bump_socks` hint at what they test, though a name in a config can be paused or finished.
- **Variant check:** load the landing page, pricing page, and checkout in three or four fresh sessions and compare. Different headlines, prices, or bumps suggest a live test or personalization, labeled Inferred. Screenshot each variant.
- **Dynamic text:** change `?keyword=`, `?utm_term=`, or `?q=` on an ad landing page and see if the headline changes to match.
- **Geo pricing:** if the browser tool supports country proxies, compare prices by country.
## Read Their Paid Search
Search results with ads, from a SERP API (DataForSEO, SerpApi) or by hand:
| Query | What it shows |
|---|---|
| Their brand name | Whether they protect it, and who bids on it |
| Your brand name | Whether they bid on you, and their ad copy against you |
| Other competitors' names | Who they treat as their rivals |
| Category terms ("best X", "X software", "X near me") | The ad copy and landing pages they pay for |
| "[their brand] alternative", "[their brand] pricing" | How they defend their own funnel |
Record the ad copy and landing page for each, dated. Hand conquesting findings to **ads** and **competitors**.
## Build the Swipe File
One file, `teardowns/<target>/funnel/swipe.md`, grouped by type, every item linked and dated:
- **Hooks:** first lines of ads, emails, and landing pages
- **Headlines and subheads**
- **Offer stacks:** bonuses, guarantees, payment plans, price anchors
- **Proof:** testimonial formats, numbers, logos, review widgets
- **CTAs and button copy**
- **Email subject lines** with their send day
Hand the swipe file to **offers**, **copywriting**, and **emails**. Copy the structure, never the words: their copy is protected.
## The Funnel Map Deliverable
`teardowns/<target>/funnel/map.md`:
1. **The diagram:** each step as a box with its price, in order, with branches for declines and abandons
2. **The offer stack table:** every product and offer, its price, its anchor, and where it appears
3. **The follow-up calendar:** day by day, email and SMS, with offers
4. **Tests seen:** tools, config names, variants
5. **Urgency check results**
6. **What to copy and what not to:** with the skill that runs each copied piece (**offers**, **cro**, **popups**, **emails**, **sms**, **pricing**, **paywalls**)
7. **Spend:** what was bought, the total, and any refunds under the published policy
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |