AI SkillsGive your Claude a marketing education.

Analyses any ad (pasted screenshot, described from memory, or linked) the way a senior creative strategist would, then turns the analysis into new ad concepts. Covers a structured teardown method (hook, awareness stage, angle, format, offer, proof, CTA), Eugene Schwartz's five stages of awareness, a taxonomy of hook types with real examples, copy devices, placement-specific format rules, a one-variable testing matrix for spinning variants from a winner, competitor research in the Meta Ad Library, TikTok Creative Center and Google Ads Transparency Center, and output templates for a written teardown and a designer-ready concept brief. Use when the user shares an ad and asks "why does this work", "tear this down", "analyse this ad", "make me ads like this", "what should my ad say", "give me hooks for X", or "what are my competitors running".

Audits and improves landing pages for conversion like a senior CRO consultant. Given a URL, screenshot or written description, it runs a structured audit (message match, above-the-fold answer test, proof, friction, speed, mobile), scores each section, ranks the top fixes by expected impact versus effort, and rewrites headlines and CTAs inline. Includes verified form design and page speed evidence, the honest math on when traffic is too low to test, and a 40-point checklist. Use when the user asks to review, audit, critique or improve a landing page, says "my landing page isn't converting", "why is my conversion rate so low", "roast my landing page", "should I redesign my site", "which button colour converts better", or "review this page" with a URL or screenshot.

Constructs and strengthens commercial offers direct-response style. Applies the value equation from Alex Hormozi's $100M Offers, builds an offer from dream outcome to named package, designs guarantees (unconditional, conditional, anti-guarantee, performance) with legal caution on results claims, applies evidence-based pricing psychology (anchoring, decoy tiers, a charm pricing reality check, annual vs monthly framing), runs the margin math on premium vs volume positioning, structures bonuses that answer objections, and tears down existing offers by scoring each value-equation variable and rewriting. Includes an offer one-pager template and a pricing test plan. Use when the user asks to create, price, improve or review an offer, pricing page or landing page, says "my offer isn't converting", "should I raise my prices", "what guarantee should I use", "how do I add urgency without being scammy", "help me package my service", or "tear down this landing page".

Designs A/B tests that can actually detect an effect and evaluates results honestly, for founders and solo marketers with real-world traffic levels. Computes required sample size from baseline rate and minimum detectable effect, sets runtime rules, catches peeking, sample ratio mismatch, multiple comparisons and segment fishing, explains what ad platform and testing tool "winners" actually mean, and delivers a ship, no-ship or keep-running verdict with a design doc and readout template. Includes a self-contained Python script for the maths. Use when the user asks about A/B testing, split testing, sample size or statistical significance, says "is my test significant", "how long should I run this test", "which variant won", "my A/B test says B is winning", "do I have enough traffic to test", "Meta says this ad won", or "should I trust this result".

Sets up and reads founder-grade marketing measurement in 2026. Covers which GA4 report answers which business question, the GA4 AI features shipped through 2026 (Analytics Advisor chat, generated insights, the AI Assistant channel for LLM referrals, custom conversion windows), UTM naming discipline, tracking ChatGPT and Perplexity traffic, honest attribution at small budgets, why platform numbers never match GA4 or actual orders, a new-site measurement checklist and a weekly readout template. Use when the user asks about analytics or measurement setup, says "set up GA4 for my site", "which channel is actually working", "Facebook says 50 sales but Shopify says 12", "how do I track ChatGPT traffic", "my traffic is all Direct or Unassigned", "what UTMs should I use", or "build my weekly marketing report".

Audits and improves a site's visibility in AI answers, ChatGPT, Perplexity, Google AI Overviews and AI Mode (AEO/GEO). Runs an evidence-based audit of extractable answers, question-shaped H2s, statistic and quote density, schema, entity clarity and third-party presence, sets up the measurement stack, and reports with honest traffic expectations grounded in published research. Use when the user asks about AEO, GEO, LLM SEO or AI search visibility, or says "how do I show up in ChatGPT", "how do I get cited by Perplexity", "we lost traffic to AI Overviews", "should I add llms.txt", "audit my site for AI search", "why does AI never mention us".

Reviews ad copy, creative descriptions and landing pages before launch and flags what will get rejected, restricted or legally challenged on Meta, Google Ads and TikTok, plus FTC exposure for US advertisers and ASA/CAP exposure in the UK. Runs a severity-graded preflight (will reject, may restrict, legal risk, fine), matches claim strength to the evidence the advertiser actually holds, and outputs a preflight report with a compliant rewrite for every flagged line. Use when the user asks to check an ad before launch, says "will this get rejected", "is this ad compliant", "my ad got disapproved, why", "can I say this in an ad", "review my landing page for ad policy", "do I need a disclaimer", or "can my influencer post this".

Plans, builds and audits ChatGPT Ads campaigns on OpenAI's Ads Manager, the ad channel launched in 2026. Covers where the platform actually stands (rollout, formats, CPC and oCPC bidding, Context Hints targeting, custom audiences, measurement), how to write Context Hints as scenario descriptions instead of keywords, when a small advertiser should or should not spend here, and produces a campaign brief. Corrects stale model knowledge, for example assuming keyword targeting exists or that Perplexity still sells ads. Use when the user asks about ChatGPT Ads, OpenAI Ads Manager, advertising inside ChatGPT, Context Hints, says "should I run ads on ChatGPT", "how do ChatGPT ads work", "set up a ChatGPT ads campaign", "what does a ChatGPT ad cost", "write my Context Hints".

Diagnoses, predicts and manages ad creative fatigue on Meta (Facebook and Instagram) and TikTok in the post-Andromeda era of 2 to 4 week creative lifespans. Distinguishes real fatigue from seasonality, auction pressure and landing page problems using numeric thresholds (frequency by funnel stage, CTR decline, CPA drift, first-time impression ratio), ranks refresh strategies by cost, plans refresh cadence by spend level, and produces a weekly per-ad fatigue report with healthy, watch, fatigued or kill verdicts. Use when the user asks about ad fatigue or creative refresh, says "my winning ad died", "my CTR is dropping", "frequency is getting high", "how often should I make new ads", "is my creative fatigued or is it the market", "my ads worked last month and now they don't", or asks for a creative fatigue check or refresh plan.

Plans, structures, diagnoses and optimizes Google Ads accounts through the 2026 AI Max transition. Covers the September 2026 auto-migration of Search campaigns to AI Max, the February 2027 Dynamic Search Ads upgrade, Performance Max controls, the Demand Gen consolidation, retired features a model might still recommend, RSA build rules, a search term and negative keyword workflow for AI Max, tCPA and tROAS ramp guidance, 2026 benchmarks and a monthly audit template. Use when the user asks about Google Ads planning, setup or performance, says "should I turn on AI Max", "my CPA is rising", "spend went up but conversions dropped", "is PMax stealing my brand traffic", "my quality scores dropped", "audit my Google Ads account", or "why did my ads stop delivering".

Plans, structures, diagnoses and scales Meta (Facebook and Instagram) ad accounts in the Andromeda and GEM era, where creative diversity has replaced audience targeting as the main lever. Covers 2026 account structure, the Advantage+ campaign structure that replaced legacy ASC and AAC, current attribution settings including incremental attribution, creative testing with numeric spend and fatigue thresholds, current benchmarks, Threads and WhatsApp placements, and diagnosis playbooks. Use when the user asks about Meta ads, Facebook ads or Instagram ads strategy, account audits or weekly reviews, says "my CPA is rising", "why did my ads stop delivering", "my ad set is stuck in learning limited", "how many creatives should I test", "should I use cost caps", or "how should I structure my Meta account".

Plans and runs TikTok ad campaigns in the post-divestiture era for a founder running their own marketing. Picks the right campaign type (Smart+, Smart+ Search, GMV Max, Spark Ads), sets per-module automation levels, structures small-budget tests around TikTok's learning phase, writes TikTok-native creative plans with refresh cadences and fatigue thresholds, uses Symphony AI tools correctly, reads GMV Max ROI correctly after the July 2026 definition change, and diagnoses delivery and conversion problems. Use when the user asks about TikTok ads, TikTok Shop ads, GMV Max, Smart+, Spark Ads, Symphony or Seedance, says "my TikTok creative is burning out fast", "my TikTok spend is not delivering", "high CTR but no conversions", "is TikTok safe to plan on after the ban", "why did my GMV Max ROI drop", or "set up a TikTok test on a small budget".