coreyhaines31/marketingskills v2.13.7

v2.13.7 — Fixes contradictions and stale facts found while rewriting marketing-skills.com's skill p…

on GitHub

Fixes contradictions and stale facts found while rewriting marketing-skills.com's skill pages:

• ab-testing (2.0.2 → 2.0.3): sample-size tables recomputed. The reference ran about 2.3× high (31,000 per variant for 3% → 3.6%) and SKILL.md about 10–20% low (12k); both now come from the two-proportion formula at α 0.05 two-sided, 80% power (14k), with the assumptions stated.
• cold-email (2.2.0 → 2.2.1): follow-up reply share aligned with the sourced Instantly 2026 benchmark (~58% of replies from email 1, ~42% from follow-ups) instead of an unsourced 55%.
• revops (2.2.1 → 2.2.2): MQL thresholds tied to their own example point tables (the two references used different weights), and SLA recipes reassign leads only under an agreed reassignment rule, matching inbound-routing.
• paywalls (2.1.0 → 2.1.1) and offers (1.0.2 → 1.0.3): paywall discount experiments now reward commitment and real deadlines (offers rules out acquisition discounts and fake scarcity); fake countdowns join the dark-patterns list; annual-prepay pricing is distinguished from acquisition discounts.
• image (2.0.2 → 2.0.3): current Black Forest Labs names (FLUX 3 Image; FLUX.2 [pro]/[flex]/[dev]/[klein]; FLUX.1 Kontext as previous generation), Midjourney V7 throughout, and an unsourced "#1 page speed killer" line removed. OpenAI's gpt-image-2.5 names checked against its deprecations page.
• directory-submissions (2.1.3 → 2.1.4): G2's deal for Capterra, GetApp, and Software Advice dated to its Jan 2026 announcement, with Software Advice no longer called a Gartner property; "AI engines penalize duplicate content" (unsourced) replaced, eval updated; a dated G2 price removed.
• aso (2.0.2 → 2.0.3): the screenshot rubric weights the first three screens instead of rewarding 8–10 (benchmarks say returns diminish past 6); the +2.5pp CPP figure is labelled as the organic-referral lift, distinct from the +5.9% average.
• analytics (2.2.1 → 2.2.2): gtag samples send GA4's recommended sign_up event, mapped from the internal signup_completed.
• programmatic-seo (2.1.0 → 2.1.1): subfolder advice rests on internal links, not a claim that subdomains split authority (its platforms reference says Google ranks neither higher); "Stats" removed from the playbook table.
• content-strategy (2.1.4 → 2.1.5): the scoring example totals 7.9, not 8.0. community-marketing (2.0.1 → 2.0.2): "5 community models" described four. copywriting (2.3.1 → 2.3.2): clarity figures credited to Founding Marketing, Ch. 3. customer-research (2.0.4 → 2.0.5): one persona minimum. copy-editing (2.2.0 → 2.2.1): its approved example dropped an em dash its own check bans. popups (2.0.0 → 2.0.1) and lead-magnets (2.1.0 → 2.1.1): duplicate related-skill entries merged.

Skill ZIP assets

Complete source catalog: 52 skills.
Claude web target: 52 skills with reviewed short descriptions.

Choose <name>-claude-web.zip for the documented single-folder web upload format.
Source ZIPs retain original metadata; they are not substitutes for omitted web ZIPs.
Live client upload, account limits, and invocation behavior have not been verified.

Web omissions (also recorded in claude-web-manifest.json):

• None.