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/team-voices.md
View on GitHub# Team Voices
Stage 4 in full: what the company's founders, leaders, and staff say in public about how they grow. Interviews and talks are often the strongest evidence a teardown gets, because people say outright what worked, when, and why. Their job posts say what they plan to do next.
## Contents
- Finding the people
- Finding what they said
- Getting the words
- Extracting claims
- Job posts as strategy documents
---
## Finding the People
Start with leaders whose work shapes growth: founders, CEO, heads of growth, marketing, product, sales, partnerships, community, and devrel. Sources:
- The team, about, and press pages, and author bylines on the blog
- X and Bluesky bios that name the company, and who the founders' accounts follow and reply to most
- Podcast guest pages and conference speaker lists
- Funding announcements and press releases, which name leaders and early hires
- LinkedIn company pages, read by hand (this repo doesn't scrape LinkedIn)
Keep it to people speaking publicly in a work capacity. Don't profile private lives, and don't contact anyone.
## Finding What They Said
| Source | How to find it |
|---|---|
| Podcasts | Search "[name] podcast" and "[company] founder interview"; Apple Podcasts and Spotify search; podcast search APIs (Listen Notes, Podscan) if the user has a key |
| YouTube talks and interviews | Search "[name]" and "[company] [conference]", sorted by views |
| Conference talks | Event sites, slides on Speaker Deck or SlideShare |
| X threads | The person's top posts, and searches for "how we grew", "lessons", "ARR", "users" from their handle |
| Blog posts and newsletters | Their own blog and guest posts, investor blogs, growth newsletters that interviewed them |
| Founder AMAs | Reddit, Hacker News, Indie Hackers, Product Hunt maker comments |
## Getting the Words
- **YouTube, TikTok, X videos:** transcripts via [ScrapeCreators](../../../tools/integrations/scrapecreators.md) (`/v1/youtube/video/transcript`, `/v1/tiktok/video/transcript`, `/v1/twitter/tweet/transcript`), or YouTube's own captions.
- **Podcasts:** many feeds publish transcripts (a `podcast:transcript` tag in the RSS feed), and many shows post episodes to YouTube too. Otherwise, download the audio and transcribe it locally.
- Save each transcript to `teardowns/<target>/voices/<date>-<person>-<show>.txt` with the source link.
## Extracting Claims
Read every transcript and pull out claims about growth into `voices.csv`:
`said_on,person,role,source,claim,topic,dated_to,quote,link`
- **Topics:** channel, tactic, number, mistake, hire, pricing, positioning, ICP, what's next.
- **dated_to** is when the thing they describe happened ("our first 1,000 users came from Reddit in early 2024"), which is often different from when they said it.
- **Quote** the exact words, short. Label every claim "reported": it's what they say, not what was measured.
- Each claim with a `dated_to` also becomes a row in `moves.csv` (evidence: Inferred, source: reported).
Then look for:
- **Claims that match the data** (they say templates drove SEO, and the template family grew before the inflection): strong evidence.
- **Claims the data doesn't back**: note them. Founders remember the story, not always the order.
- **Things they say didn't work**: these go straight to plays to skip.
- **Repeated phrases** across interviews: that's the positioning they've rehearsed.
## Job Posts as Strategy Documents
Read the full text of current and past job posts, not just titles. Start with `teardown.mjs jobs <domain> --out teardowns/<target>/jobs.csv`, which finds a public Greenhouse, Lever, or Ashby board and summarizes it: roles by function, marketing channels being hired for, tools named, languages required, and "first marketing hire" wording. Past posts are on Wayback (`web.archive.org/web/*/target.com/careers*`) and on job boards. Saving `jobs.csv` each month lets Watch mode spot new roles by id.
What to pull from each post:
- **The mandate:** "own our PLG motion", "scale programmatic SEO to 10,000 pages", "launch our partner program", "build our first outbound team". This is the plan, written down.
- **Numbers:** team size, targets, budgets ("manage $2M/year in paid").
- **Tools named:** the stack the team uses.
- **Reports to:** where the function sits.
- **First seen:** the date the post first appeared, which dates the push.
Add each mandate to `moves.csv` as `kind: hire` with the posting date. [TheirStack](../../../tools/integrations/theirstack.md) searches job posts by company if the user has it.
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: 323k | 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: 323k | 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: 323k | Updated | Claude CodeCodex | Verified | Install |