Marketing Attribution Guide
Which marketing actually drives revenue — models, blind spots, and first-party truth
- What
- Which marketing actually drives revenue — models, blind spots, and first-party truth
- Cost
- Free
- Needs
- none — guidance for marketing teams and founders
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — answers the hardest marketing question: which marketing actually drives conversions and revenue. Pillar A is interpretation: what attribution can and can't tell you, the models (first-touch, last-touch, multi-touch), the three measurement paradigms, self-reported attribution, reconciling conflicting sources ("Google/Meta says X but GA says Y"), and the blind spots — with a fork by business type. Pillar B is owning your attribution: the identity graph, closing the `identify()` gap, stitching conversions on a third-party domain, guardrails not to skip, and reporting the last mile. Produces a structured Attribution Readout (the question, source of truth, what each source says, model comparison, confidence and gaps, recommendation) with tool integrations and related skills. Distinct from influencer-marketing (lot 32, partnerships) — this one is measurement truth. Pure guidance, no paid services. By @coreyhaines31, listed here with credit to its creator. Update (October 2026): this listing's install prompt was suspended on October 20, 2026 by an automated scan that matched the words "cookies" and "webhook" in the guide; a human review found no secret reading and no data sent by the skill — the only flow it teaches passes a site's own anonymous analytics id to a tool the site owner chose and back to the owner's own handler, under a fail-closed guard against personal data — and the suspension was lifted on October 21, 2026 with a regenerated prompt at upstream version 1.1.6 that discloses all of this. Skill Harbor never reviews the code, review it yourself before use.
Version: 1.1.6
Quick answer
Marketing Attribution Guide is a free Muse skill answering which marketing actually drives conversions and revenue. Pillar A is interpretation: what attribution can and cannot tell you, the models, the three measurement paradigms, self-reported attribution, reconciling conflicting sources, and the blind spots, with a fork by business type. Pillar B is owning your attribution, from identity graph to reporting.
Install
Copy the install package below, then paste it into Muse🧪 Install ran in our test sandbox · Nov 18, 2026📌 Pinned version: commit e85f3cbStep 0 — before installing anything: read this install prompt, summarize in a few lines what this build does and what installing it involves, then ask me to confirm before you proceed. Prerequisites: none — guidance for marketing teams and founders Install "Marketing Attribution Guide" for me. Which marketing actually drives revenue: attribution models and their blind spots, reconciling dashboards that disagree ("Google/Meta says X but GA says Y"), and owning your first-party attribution — the identity graph, closing the identify() gap, stitching conversions that complete on a third-party domain, and the guardrails not to skip. It produces a structured Attribution Readout. Repository: https://raw.githubusercontent.com/coreyhaines31/marketingskills/e85f3cbc2e4a79bf33b2a12e9bdd9087d414735b/skills/attribution/SKILL.md (the skill is the skills/attribution/ folder at that commit: SKILL.md, references/ with four files, and evals/evals.json) Pinned version: install the code at commit e85f3cbc2e4a79bf33b2a12e9bdd9087d414735b exactly (Skill Harbor tested this version). Do not substitute a newer version. If this commit is unreachable, stop and tell me. [FINGERPRINTS] Files at the pinned commit, from Skill Harbor's review of 2026-10-21: - SKILL.md: SHA-256 950ed42f471fa5484eae4c122a2c2cd54aa4916ce65581b5cb2f51fe8c630710, 23,423 bytes, 23,284 characters - references/attribution-models.md: SHA-256 503c61bf2f38d54755398fc5b1bb0d190e989c59494b097407652ae65bf474af, 6,783 bytes, 6,727 characters - references/by-business-type.md: SHA-256 224a937e09842dbd7a8e00fb67f060cd1ac50f58f9f576be8d015a7c800c9de4, 7,857 bytes, 7,806 characters - references/first-party-tracking.md: SHA-256 a8e12b00d3d049d4b32e3a2785e504936ee7ae33a523a2a1717c338bd4140b6c, 19,867 bytes, 19,630 characters - references/measurement-paradigms.md: SHA-256 5efca610abaa3a80a31cd6be5e835e4449d54462c5314e97b055b86413dee5b2, 10,598 bytes, 10,511 characters - evals/evals.json: SHA-256 795a908eacac5cd2213d03d635fd92d94a6d7d66544b68273a0dbdcc2609607b, 13,186 bytes, 13,168 characters If any of these does not match, it changed since the review: do not install it yet, tell me what you found and let me decide. Updated October 21, 2026 — please read this disclosure before installing; it is part of the prompt on purpose. Skill Harbor regenerated this install prompt at upstream version 1.1.6, and this listing carries a suspension history that we disclose rather than hide. On October 20, 2026, an automated scan suspended this listing's install prompt under the pattern name "exfiltration". The scan pairs a "secret source" line and a "send channel" line found anywhere in the same file: here, the word "cookies" in a sentence about measurement limits ("cookies expire, sessions fragment…" — why attribution data is incomplete) and the word "webhook" 117 lines later, in a different section. There is no flow between those two lines, and the file contains no code that reads any secret: no cat, curl, wget, requests or fetch calls, no .ssh, .env, password or API key anywhere in it; its only code blocks are the Markdown template of the report it produces. The one data flow the guide actually teaches is described plainly in the safety check below: a site's own anonymous analytics id passed to a third-party tool the site owner chose, and returned by that tool's webhook to the owner's own handler, under an explicit fail-closed guard that forbids passing personal data. The same passage, word for word, had already triggered the same suspension on October 14, 2026, and Skill Harbor's owner lifted that one as a false positive after human review. What changed upstream in version 1.1.6 is only the version number and three cross-references (the sibling skill "ads" is now "conversion-tracking"); the flagged passage is unchanged. After a fresh human examination of the full file, Skill Harbor's owner decided on October 21, 2026 to lift the suspension through this regenerated prompt, which discloses all of the above. Because the words "cookies" and "webhook" remain in the guide — they are ordinary vocabulary for what it teaches — an automated scan may flag this listing again; the human review conclusion stands in this prompt: no secret is read, and nothing is sent anywhere by the skill itself. 1. Fetch the whole skills/attribution/ folder at the pinned commit into a temporary folder, then compare SKILL.md and all five other files with the fingerprints from Skill Harbor's review of 2026-10-21 (see [FINGERPRINTS] above). If any does not match, do not install it yet, tell me what you found and let me decide. 2. Safety check: read SKILL.md and the references/ files and review them for anything suspicious (unexpected network calls, shell commands, credential harvesting); STOP on any red flag and tell me. Tell me in plain words what the skill does when used, which services it connects to, whether it reads or sends any personal information or keys, and anything that seems unusual. Confirm these points from the files: - Guidance only: six files — Markdown guides plus one JSON file of evaluation cases — no scripts, no executable code, no background service, and nothing runs on its own. Nothing is sent anywhere and nothing is executed during installation itself: installing only downloads the pinned folder from GitHub and copies it. - The one data flow the guide teaches, stated plainly: when a conversion completes on a third-party domain I do not control (a booking tool, a hosted checkout — of my choosing), a capture-phase link decorator appends my visitor's anonymous analytics distinct_id to the outbound URL via the tool's metadata passthrough; the tool stores that metadata and returns it in its webhook to my own webhook handler, which fires an identity merge and a conversion event in my own analytics. The anonymity guard is fail closed: only ever the anonymous id — ids that look like personal data (containing @) are rejected, length is capped, and when identity is ambiguous, nothing is sent. - This listing's suspension history is real and is disclosed in the update note above, not hidden: an automated scan matched the words "cookies" (measurement limits) and "webhook" (the return channel just described) in the same file and called it "exfiltration". Human review of the full file found no secret reading and no sending by the skill itself; the only flow is the one in the previous point, under the fail-closed guard. Skill Harbor's owner lifted the suspension on October 21, 2026. - No accounts, API keys or credentials are needed or involved, and the repo contains no secrets in code. The analytics tools the guide discusses are my own, configured by me, outside this install. 3. When I say "yes, install", install it as a skill: copy the whole fetched skills/attribution/ folder (SKILL.md, references/ and evals/, keeping the subfolders so the relative links in SKILL.md keep working) into the agent's skills directory, in a folder named "coreyhaines31-attribution". 4. Verify with no network calls: frontmatter valid, files in place, and the copied SKILL.md still matches its fingerprint above. 5. Report what was installed, where, and what I still need to do myself (nothing to configure for the install itself — the guide is used in conversation; applying its first-party tracking pillar later is work on my own site and analytics setup, under my own control). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me. Never print or paste a secret value.
Saved to your recent installs. Find it anytime on /connect.
🔒 Security, provenance and what this touches
- It can contact outside services over the network.
- It works with an API key or a token that you provide.
What the scan found — excerpts from the install prompt
- makes network calls — Updated October 21, 2026 — please read this disclosure before installing; it is part of the prompt on purpose. · - The one data flow the guide teaches, stated plainly: when a conversion completes on a third-party domain I d · - This listing's suspension history is real and is disclosed in the update note above, not hidden: an automate · Mis à jour le 21 octobre 2026 — lis cette divulgation avant d'installer ; elle fait partie du prompt à dessein · - Le seul flux de données que le guide enseigne, dit clairement : quand une conversion se conclut sur un domai · - L'historique de suspension de cette fiche est réel et il est divulgué dans la note de mise à jour ci-dessus,
- uses API keys or tokens — Updated October 21, 2026 — please read this disclosure before installing; it is part of the prompt on purpose. · - Aucun compte, clé API ni identifiant n'est requis ni en jeu, et le dépôt ne contient aucun secret dans le co
The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?
Verification certificate
Dated facts about what was checked for this listing, and how. This is not a safety certification.
- Pinned version
- e85f3cb · 2026-10-21
- Sandbox install test
- Install ran in our test sandbox · Nov 18, 2026
- Creator identity
- Not verified
- Fingerprint watch
- Last checked · Nov 19, 2026 : no change detected in the files linked by this install prompt
- Install prompt analysis
- 2026-10-21 : makes network calls · uses API keys or tokens
Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify
Questions
How do I install a skill?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the skill up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a skill before installing it.