What is REM (Reverse Engineer Marketing)?
REM takes a domain or an App Store or Google Play link and collects what a company's marketing leaves in public: tracking pixels (including those inside Google Tag Manager containers), sitemaps and programmatic page families, Wayback Machine snapshots of the homepage, app store data, Hacker News launches and, with a paid Scrape Creators key, Meta, Google and LinkedIn ad libraries, social accounts, founders and Reddit threads. Each item becomes numbered evidence, and a model writes a report and a copy-this playbook that must cite it.
It runs as a command line tool, as an MCP server with 7 tools for Claude Code, Codex or Cursor, and as a local web app on 127.0.0.1:4321 that writes the report through your logged-in Claude Code or Codex CLI. A validator checks that citations resolve and that no ad spend is invented. A diff command compares two runs of the same company.
The repository is MIT licensed and comes from DansUGC, a company selling user-generated video clips for ads.
Is REM (Reverse Engineer Marketing) worth it?
REM's strongest choice is its evidence rule. Every fact in a report must cite a numbered observation or calculation, anything it could not check is listed as unknown rather than absent, and a tracking pixel is reported as intent to advertise, not as spend. Competitive teardowns written by a model usually blur those lines; the bundled validator is built to catch a report that does.
The free layer is useful on its own: pixels inside Tag Manager containers, sitemap folders that reveal comparison and integration page farms, the AI crawlers a site blocks, and homepage positioning over time from the Wayback Machine. The paid layer is cheap. A full run takes about 25 to 50 Scrape Creators credits; at the $47 pack of 25,000 credits that is under 10 cents, and responses are cached for 24 hours.
The catch: as of 14 November 2026 the project is six commits made on a single day, 18 October 2026, with nothing since. The README's npx reverse-engineer-marketing commands fail because no package of that name is on npm; you have to clone, install and build it yourself on Node 22 or newer. One of those six commits added DansUGC sales links with tracking codes to the README, the web app, the CLI and the generated reports, so the teardowns you share carry someone else's advertising. The coverage gaps are honest but real: Apple Search Ads only for the EU, no public library for TikTok or X ads, and ad counts taken from the first pages of each library.
Our take: a marketer who can run a Node project gets a cheap, well-reasoned first pass on a competitor in minutes. Until a package ships and a second week of commits appears, it is a method worth borrowing more than a tool to depend on.
Where it earns its place
A founder of a three-person consumer app startup
Find out which ads a bigger rival has kept running for months
Build from source, add a Scrape Creators key and run it on the rival's domain; the report leads with the longest-running Meta ads, their hooks and landing pages. About 25 to 50 credits a run, a few cents.
A B2B content lead planning next quarter's SEO
See whether a competitor's traffic comes from programmatic comparison and integration pages
Run with --no-ai and --only seo for a free pass over sitemaps and page families, then read evidence.json. No keys and no model call.
An agency strategist tracking four clients' competitors
Report each month what changed in a rival's advertising
Connect the MCP server to Claude Code and use the what_changed prompt, which re-runs a company and diffs new and dropped ads, follower growth and repositioning against the last run.
Before you commit
- Setup
- Clone the repository, npm install and build on Node 22 or newer; the published npx command does not work because the package is not on npm. The demo command opens a recorded teardown without keys.
- Integrations
- Claude Code, Codex, Cursor or any MCP client; Claude Code or Codex CLI as the report writer; Anthropic API optional; Scrape Creators for ad libraries, social and Reddit data.
- Data and privacy
- Reads public pages and public ad transparency libraries only. The web app listens on 127.0.0.1, stores the Scrape Creators key in ~/.config with mode 600 and caches responses for 24 hours.
- Cost at scale
- About 25 to 50 Scrape Creators credits a run, so 100 competitor runs fit in the $47 pack of 25,000 credits; report writing uses your existing Claude or ChatGPT subscription.
- Limits
- Apple Search Ads visible for the EU only; TikTok and X ads have no public library; Reddit's ad library is closed; ad counts are samples; founder and TikTok matches can be wrong.
What’s good
- Every claim in a report cites numbered evidence; unchecked items are listed as unknown
- Useful free layer: pixels inside Tag Manager, sitemap page families, Wayback history
- A full paid run costs under 10 cents in Scrape Creators credits
- Report writing runs through your Claude Code or Codex plan, or skips AI entirely
- MCP server with 7 tools and a validator for agent-written reports
What’s not
- npx install in the README fails: the package is not on npm
- Six commits on one day and no activity since
- Reports and the web app carry DansUGC sales links with tracking codes
- Apple Search Ads covered for the EU only; no TikTok or X ad library
Best for
Growth marketers and founders who can run a Node project and want a cited first read of a competitor's channels before paying for a research tool.
REM (Reverse Engineer Marketing) alternatives
REM (Reverse Engineer Marketing): common questions
- Is REM (Reverse Engineer Marketing) worth it?
- REM has a sound method for reading a rival's marketing from public traces, wrapped in a one-day code drop whose advertised install command does not work. We rate it 3.3 out of 5.
- How much does REM (Reverse Engineer Marketing) cost?
- Free under MIT. The free sources need no key. Ad libraries, social and Reddit data use Scrape Creators at about 25 to 50 credits a run (100 free credits; packs from $47 for 25,000). The report runs on your Claude or ChatGPT plan, or on the Anthropic API per token.
- What are the downsides of REM (Reverse Engineer Marketing)?
- npx install in the README fails: the package is not on npm; six commits on one day and no activity since; reports and the web app carry DansUGC sales links with tracking codes; apple Search Ads covered for the EU only; no TikTok or X ad library.
- Who is REM (Reverse Engineer Marketing) best for?
- Growth marketers and founders who can run a Node project and want a cited first read of a competitor's channels before paying for a research tool.
- How much does REM (Reverse Engineer Marketing) cost for a team or in production?
- About 25 to 50 Scrape Creators credits a run, so 100 competitor runs fit in the $47 pack of 25,000 credits; report writing uses your existing Claude or ChatGPT subscription.
- Is REM (Reverse Engineer Marketing) safe for company data?
- Reads public pages and public ad transparency libraries only. The web app listens on 127.0.0.1, stores the Scrape Creators key in ~/.config with mode 600 and caches responses for 24 hours.
- What are the best REM (Reverse Engineer Marketing) alternatives?
- The closest alternatives we have reviewed are OpenSEO.
Answers checked Nov 14, 2026.