The Marketing Skills LibraryFree AI skills for Claude & ChatGPT

Builds B2B positioning from customer switching evidence — alternatives, unique attributes, value, best-fit customers, category — scored 0–4, with a swap-tested statement and the riskiest assumptions.

Turns positioning and proof into a messaging house — category POV, value proposition, three proof-bound pillars — plus a persona-by-stage message map, proof gaps, and words to use and avoid.

Designs 5-second, comprehension and preference tests with neutral questions and sample math, then scores real responses 0–4 on four dimensions and calls a winner only when the data supports one.

Maps competitors' public claims from dated homepages, ads, pricing pages and reviews into a territory matrix — crowded, contested, whitespace — and scores 3–5 territories you could own.

Turns reviews, transcripts, surveys and tickets into themes ranked by frequency × intensity, graded for confidence, plus a verbatim message bank where every quote traces to its source.

Tiers a B2B launch with an anchored five-criterion rubric, then builds the tier's asset plan with owners and dates, an enablement-first timeline, a go/no-go gate and baseline-based metrics.

Writes a one-page B2B campaign brief: one objective, an evidence-tagged insight, a single-minded proposition, justified channels, a pre-registered primary metric and kill criteria set before launch.

Writes Google responsive search ads and LinkedIn single-image ads with a printed character count on every line, one angle per variant from a fixed taxonomy, proof-cited claims and a test plan.

Scores a B2B landing page on eight anchored dimensions with the page quoted behind every finding, counts unsupported claims, and ranks fixes by impact × effort.

Turns a target query and real SERP results into a writer-ready brief: intent verdict with evidence, an information-gain plan with SME questions, an answer-first outline and counted titles.

Assigns every URL in a Search Console export one decision (refresh, consolidate, redirect, prune, noindex or leave) with a reason, flags cannibalization, and protects pages that carry value.

Builds a repeatable buyer prompt panel, scores how AI assistants mention, recommend, cite and describe your brand, and turns wrong answers into page fixes and PR targets.

Turns a customer interview into a case study where every metric traces to a timestamped quote or data row, in three formats, plus the approval checklist your customer signs.

Turns an executive's notes, voice memos or talks into three LinkedIn posts on one idea, in their voice, each built on a first-hand specific from the input and nothing invented.

Builds behavior-triggered B2B nurture sequences where every email has a send and exit condition, one job or objection, sourced proof and one CTA, measured on clicks, replies and conversions.

Builds a fit + intent lead scoring model with every weight justified by lift on your own won and lost leads, a back-tested MQL threshold, and the SLA and governance to keep it honest.

Builds a controlled UTM vocabulary with a fixed medium list and campaign pattern, then validates pasted tags with fixed error codes and returns corrected, CSV-ready URLs.

Reconciles ad-platform, analytics, CRM and "How did you hear about us?" data into one conversion count, models side by side, and a High/Medium/Low confidence rating per channel.

Turns spend and pipeline by channel into pipeline per dollar and marginal returns with the math shown, a capped reallocation with a test reserve, and three scenarios that name their assumptions.

Turns a quarter's marketing metrics into a board update that leads with the miss: three things, a defined scoreboard, variance by driver with the math, a 6–8 slide outline and a one-page memo.