marketing-skills

Build a polished, hosted Instagram Insights dashboard for any Instagram account using the PorterMetrics MCP — a 4-page report (Overview, Posts, Stories, Audience) that ships as a live shareable URL. Use this skill WHENEVER the user wants an Instagram / IG analytics dashboard, report, or "insights" page built with Porter / PorterMetrics — even if they only say "make me an Instagram report", "dashboard for my IG account", "replicate this Instagram dashboard", "Porter report for Instagram", or hand you a Porter Instagram report URL to reproduce. It offers a one-call CLONE path that works anywhere (including plain chat, no build tools) plus a full build path, so every run produces a near-identical dashboard. Do NOT hand-roll an Instagram report from scratch when this skill applies.

Build a complete, hosted marketing dashboard on Porter Metrics end-to-end — from intake interview and SUMAS planning, through a brand-approved design kit (Porter brand or white-label extracted from any client website), to a live published report URL with real data. Use when a user with the Porter MCP asks for "a dashboard", "a client report I can share", "a live report for account X", or wants to white-label a report with a client's brand. Connector- agnostic (Google Ads, Meta, GA4, ...) and framework-agnostic (Next.js template is the default path, any static-site stack meeting the Porter contract works).

Track each keyword's Google Ads Quality Score over time and flag the keywords whose QS is DECLINING — the quiet erosion that raises your CPC and drops your rank without any alert firing — then attribute the drop to the component responsible (Expected CTR, Ad Relevance, or Landing Page Experience). Use this skill whenever the user asks about Quality Score trend, QS dropping / falling over time, "why did my Quality Score go down", CPC creeping up for no obvious reason, keywords slipping in rank, or which component (expected CTR / ad relevance / landing page experience) is hurting a keyword — even if they don't say "Quality Score". This skill diagnoses WHERE QS is eroding and WHY (which component) ONLY; the ad-copy fix belongs to the complementary `rsa-strength-copy-diversity-audit` skill, the landing-page fix to `landing-page-cro-audit`, and the point-in-time keyword↔ad↔landing relevance snapshot to `keyword-ad-landing-alignment`.

Quantify the TOTAL wasted spend in a Google Ads account and pinpoint the biggest leaks — money spent on search terms, keywords, ad groups, or campaigns that returned ~0 conversions or cost far more per conversion than target. Rolls the leaks up to one dollar number and a % of spend, then lists the top leaks by scope, each with the dollars wasted and a recommended action. Use this skill whenever the user asks how much they're wasting, where the money is leaking, "what spend is doing nothing", zero-conversion keywords/terms, spend with no return, or an account-level waste roll-up — even if they don't say "wasted spend". Quantifies and prioritizes waste ONLY; writing the negatives belongs to `negative-keywords`, reallocating the freed budget to `spend-allocation`, and classifying WHY a term is irrelevant to `search-term-classifier`.

Give the advertiser the honest Performance Max read — is each PMax campaign earning its budget against the account's Search baseline (CPA/ROAS), what asset-group-level performance is actually visible, and — explicitly — what Google's API does NOT expose for PMax (search terms, placements, audience detail), so the reader trusts the numbers that ARE real. Use this skill whenever the user asks about Performance Max, PMax, "is my PMax working / worth it", PMax spend, PMax vs Search, asset groups, or the PMax black box — even if they don't say "diagnostics". Judges PMax earning-vs-baseline and the visible asset-group signal ONLY; full search-term mining and placement control are API-limited (named as gaps, not faked), and bid targets belong to the complementary `bid-strategy` skill.

Build an exhaustive, human-readable Google Ads CHANGE HISTORY / changelog for a Porter-connected account — what changed, when, by whom, translated into plain English (the exact negative keyword, the budget before→after, the tCPA/tROAS, the paused campaign), grouped by day and change type, and overlaid against the daily CPA so you can see which performance moves followed which edits. Use when the user asks "what changed in my Google Ads account", "change history", "changelog", "account audit log", "who changed what", "qué cambios se hicieron en la cuenta", "historial de cambios", or wants to correlate account edits with CPA/performance. Runs on the Porter Metrics MCP (google-ads connector). One account at a time. Google retains only the last 30 days of change history.

Tell the advertiser whether each Google Ads campaign is on track to spend its budget over the period — over-pacing (burning too fast, hitting the daily cap and losing impressions to budget), under-pacing (won't spend what's allocated), or on-track — with a projected end-of-period spend per campaign. Use this skill whenever the user asks about budget pacing, spend pace, "am I on track to spend my budget", over-pacing / under-pacing, burning budget too fast, campaigns hitting their daily cap by mid-day, under-delivering budgets, or month-end spend projection — even if they don't say "pacing". Judges spend PACE over time ONLY; where to reallocate budget belongs to the complementary `spend-allocation` skill, and bid targets to `bid-strategy`.

Análisis completo de competidor en Meta Ads a partir del solo nombre. Extrae ads activos de la Biblioteca de Anuncios de Meta para un brand (videos E imágenes), deduplica por SHA256, transcribe audio, extrae frames, hace OCR + análisis visual, y arma un REPORTE VISUAL Porter (HTML autocontenido). Activar con /meta-ads-research o cuando el usuario pida "scrape ads de meta de X", "research creativo de Y", "trae los ads activos de Z", "auditar competidor en Meta", "analiza al competidor X" o similar. Output: JSON canónico + brand_brief.md + reporte HTML estilo Porter. Usa Apify + Deepgram + Anthropic vision + PaddleOCR + ffmpeg.

The guided wizard/checklist that walks a (non-technical) user through EVERY decision needed to launch a complete, correctly-configured Meta Ads campaign — naming, objective, bidding, budget, targeting, placements, promoted object, creative, copy, tracking — validating each answer against the account and Meta's rules, explaining the options and trade-offs in plain language, then producing a validated "campaign spec", showing a summary for the user's OK, and creating everything PAUSED via the executor skills. Use whenever the user wants to launch/set up a Meta campaign but isn't sure what to choose, asks "help me set up a campaign", "what do I need to run ads", "guide me", or when a campaign request lacks the required parameters. This is the BRAIN; campaign-setup / adset-setup / asset-upload / ad-setup are the executors it dispatches.

Create the AD (the creative + copy + CTA + destination) inside a Meta ad set — routing by creative type (single image, video, carousel, dynamic creative, per-placement multi-format), wiring the copy, the CTA button, the destination link with UTM tracking, and the lead form for LEADS. Use whenever the user wants to build/create a Meta ad, attach a creative to an ad set, write the ad copy/headline/CTA, or launch the actual advertisement, even if they don't say "Porter". Scope: the ad object only; uploading the asset is meta-ads-asset-upload, targeting/placements is meta-ads-adset-setup.

Create and configure a Meta (Facebook/Instagram) Ads AD SET under an existing campaign — optimization goal, audience/targeting (geo, age, interests, custom/lookalike audiences), placements, promoted object (page/pixel/lead form), schedule, ad-set budget & bid (only when the campaign is not CBO), always PAUSED. Use whenever the user wants to define WHO sees the ads and WHERE, set targeting/placements/audience, or add an ad set to a Meta campaign, even if they don't say "Porter". Scope: the ad-set level only; the campaign shell belongs to meta-ads-campaign-setup and the ad/creative to meta-ads-ad-setup.

Get a creative asset (image or video) INTO a Meta ad account and return the image_hash / video_id that an ad needs — sourcing the file from Google Drive (via the Porter storage/google-drive connector) or a public URL. Use whenever the user wants to upload an image/video from Drive to Meta/Facebook Ads, "subir una imagen de Drive a Meta", turn a Drive asset into an ad creative, or prepare creatives for an ad. Scope: the ingest step only (file → hash/id); assembling the ad (copy, CTA, placements, formats) belongs to meta-ads-ad-setup.

Create and configure a Meta (Facebook/Instagram) Ads campaign object on any Porter-connected ad account — objective, budget strategy (CBO), bid strategy, special ad categories, naming — always created PAUSED for safety. Use this skill whenever the user wants to launch, create, set up, or spin up a Meta/Facebook/Instagram campaign, or says "crea una campaña", "arma una campaña en Meta", "set up a campaign", "launch a new campaign", even if they don't say "Porter". Scope: builds ONLY the campaign shell + its budget/objective settings; the ad set (targeting, schedule, adset budget) and the ad/creative belong to the complementary meta-ads-adset-setup and meta-ads-ad-setup skills.

End-to-end orchestrator — take a creative sitting in Google Drive and turn it into a ready-to-launch (PAUSED) Meta/Facebook ad, in one guided flow. Use when the user says "sube este creativo de Drive a Meta y ármame el anuncio", "from Google Drive to Meta Ads", "publica (en pausa) este video/imagen de mi Drive", or wants the whole chain (asset → campaign → ad set → ad) from a Drive file without stitching the sub-skills by hand. It chains meta-ads-asset-upload → meta-ads-campaign-setup → meta-ads-adset-setup → meta-ads-ad-setup, adds the Drive-source resolution + transport decision that make it CONSISTENT, verifies, and (for tests) tears down. Everything is created PAUSED; activation is a human decision.

Duplicate the best-performing Meta (Facebook/Instagram) campaign / ad set / ad. Meta has NO native "duplicate" via the Porter MCP, so this reads the winner's full structure and RECREATES it PAUSED using the *_create actions. Use when the user wants to clone or scale a proven winner, duplicate a campaign/ad set, replicate a top performer into a new audience or budget, or A/B a variant. The winner = best by PERFORMANCE (CPA/ROAS with enough volume), NOT the one that spent the most.

Read and interpret Meta (Facebook/Instagram) Ads performance — spend, impressions, reach, clicks, CTR, CPC, CPM, frequency, conversions/actions, CPA/ROAS — at account / campaign / ad-set / ad level, with breakdowns (age, gender, placement, region, device), to answer "how is it doing and WHERE does it leak". Use whenever the user asks for results, performance, what's working/not, where budget leaks, or to feed an optimization decision. Read-only. Scope: reading + diagnosing performance; changing budgets/status belongs to the setup skills / meta-ads-rules-optimizer.

Optimize Meta (Facebook/Instagram) ad performance by rules — read results, then pause / scale / adjust ad sets and ads by CPA/ROAS thresholds (e.g. pause ad sets over 3× target CPA, scale winners gradually, cut fatigued creative). Use when the user wants to optimize, prune losers, scale winners, cut wasted spend, or set up "if CPA > X then pause" logic. NOTE: Meta's native automated rules are NOT exposed by the Porter MCP, so this runs as ASSISTANT logic (read insights → decide → write). Always confirm each write with the user; scale gradually and respect the account-safety limits.

Create and manage Meta custom audiences and lookalikes on any Porter-connected ad account — create a seed audience, upload your customer/lead data (email, phone, etc., hashed), poll until ready, build a lookalike to scale, and list/read existing audiences for retargeting. Use whenever the user wants to build a custom audience, upload a customer list, create a lookalike, set up retargeting, or see what audiences exist, even if they don't say "Porter". Scope: audiences only; using them in an ad set (targeting) belongs to meta-ads-adset-setup.

Design and generate a consistent naming convention for Meta (Facebook/Instagram) Ads across campaign, ad set, and ad levels — plus the matching UTM/URL tracking — on any ad account. Built on the SUMAS levers (producto, objetivo, ángulo, audiencia…): each lever gets ONE home level, names are lowercase and machine-parseable, and the UTMs build themselves from the names via Meta dynamic params. Use whenever the user wants to plan how to name their campaigns/ad sets/ads, set up a naming convention, structure a Meta account, define UTMs, or asks "cómo nombrar mis campañas", "convención de nombres", "nomenclatura", "estructura de campañas", "naming convention", "how should I name my ads", "set up UTMs". Account-agnostic: ships generic defaults, reads a per-account spec the user provides. It PLANS and GENERATES names/UTMs — it does not create objects in Meta (that's meta-ads-campaign-setup / adset-setup / ad-setup).

Keep a Meta (Facebook/Instagram) ad account healthy and avoid getting it banned/disabled/restricted while running ads or automations, and connect accounts the SAFE way. Use whenever a user connects a Meta account, worries about bans/restrictions/"unusual activity", asks about API rate limits or automation limits, or before ANY bulk/automated write on Meta Ads. Core rule: operate ONLY through Meta's official Marketing API (via the Porter MCP) — never browser automation, scraping, or unofficial/direct-app access.

Classify Meta/Facebook ad creatives as Unicorn / Winning / Steady / Fatiguing / Losing / Testing using an industry-grounded creative-strategy framework (3-stage funnel + level-vs-benchmark + weekly fatigue trend). Sibling of performance_decay, tuned for ad-rate metrics instead of SEO traffic. Use when a marketer asks which ads to scale, refresh, or cut.

Recreate the golden multi-page Meta (Facebook/Instagram) Ads dashboard via the Porter Metrics MCP — Overview, Campaigns (SUMAS), Creatives (with Unicorn/Winning/Losing verdicts + retention + tooltips), Budget pacing, Funnel, Audience placements, Conversion windows. Use when a user with the Porter MCP and a connected Meta Ads account wants a complete, hosted Meta Ads reporting dashboard.

Create and manage Meta Instant lead-generation forms on a Facebook Page, and retrieve the leads submitted — for OUTCOME_LEADS campaigns. Build the form (questions, privacy policy, thank-you screen), list forms on a Page, and pull the collected contacts. Use whenever the user wants a lead form, an instant form, to capture leads on Facebook/Instagram, or to download the leads, even if they don't say "Porter". Scope: the lead form + leads; wiring it into an ad is meta-ads-ad-setup (lead_gen_form_id), and the campaign/ad set setup is the campaign/adset skills.

For each Google Ads ad group, judge whether its Responsive Search Ads are built to win the auction — strong Ad Strength, diverse non-redundant headlines, and pinning that doesn't strangle rotation. Use this skill whenever the user audits a Google Ads account, mentions Ad Strength, responsive search ads, RSAs, ad copy, headlines, pinned headlines, duplicate / near-identical headlines, "why is my Ad Strength Poor", or asks which ads to rewrite — even if they don't say "ad copy". Judges the AD COPY ITSELF (strength + headlines + pinning) ONLY; ad EXTENSIONS/assets belong to the complementary `ad-extensions-assets-audit` skill, and message-match to the keyword/landing belongs to the `keyword-ad-landing` cluster.

Scan Google Ads accounts for the TWO creative problems that actually waste money — ads pointing to broken/dead landing pages, and disapproved (rejected) ads/assets that run handicapped. Runs across one account or many and returns a prioritized list of what to fix. Use this whenever the user asks to check ad/landing health, find broken links or 404s in ads, find disapproved/rejected ads, audit creative for wasted spend, or "what's broken in my Google Ads". Deliberately ignores Ad-Strength hygiene (headline counts, char usage, pinning) — those don't reliably move performance (see metric-levers.md). One account at a time; built to loop across an account list.

Map a Google Ads account's SEARCH creative into a clean, structured graph — every ad (id + name) and below it its variations (headlines and descriptions), each with its pin position, character length, Google performance label, and its own impressions/clicks/conversions/cost, plus the extensions linked to it. This is the PRODUCER of the per-account `creative_graph` that every other creative skill (performance, copy-hygiene, rewrite) consumes. Use this skill whenever the user wants to inventory / structure / map their ad copy, asks "what headlines and descriptions am I running", needs the relationship between ads and their assets, or is starting any creative audit — even if they don't say "inventory". Maps and structures ONLY; it does NOT judge performance, repetition, or copy quality (those are the consumer skills). Search/RSA first; PMax & other types are documented in references/campaign-types.md.

Compute the cost per INDIVIDUAL conversion action — cost per phone call, cost per form submit, cost per booking — which Google Ads does not give natively (it only shows one blended cost/conversion across all primary actions). Splits conversion counts by campaign × action, divides by spend, and flags when a campaign's real driver doesn't match its name/goal. Use this skill whenever the user asks "what's my cost per phone call / per lead form / per [action]", per-conversion-action CPA, cost by conversion type, or which action a campaign actually drives — even if they don't say "CPA". Reports a FULLY-LOADED per-action CPA (never additive); the conversion setup verdict belongs to `conversion-tracking` and budget moves to `spend-allocation`.

Audit whether a Google Ads account's ad URLs carry consistent UTM tags so GA4 and the CRM can attribute each click back to the campaign / keyword that drove it. Checks presence, scheme consistency, and dynamic utm_term across campaigns, then recommends a consistent UTM scheme + auto-tagging. Use this skill whenever the user audits a Google Ads account, asks about UTM tags, URL tagging, campaign tracking, attribution, "can my CRM see which campaign drove this lead", auto-tagging / GCLID, lost source data, or untagged ad links — even if they don't say "UTM". Judges URL TAGGING HYGIENE only; conversion-action setup belongs to `conversion-tracking`, and GA4 configuration is a different connector.

Assemble a complete Google Ads account audit as a single scrolling HTML narrative document. Orchestrates all Google Ads analysis skills, orders findings by dollar impact, and renders them as a professional audit report with Executive Summary, KPI Scorecards, per-section findings tables + callouts, What's Set Up Right, and a prioritized Action Plan. Use this skill when the user asks to audit, review, or health-check a Google Ads account. Output is a Porter report (single HTML page), not a dashboard.

Classify every Google Ads search term as brand (own brand + misspellings), competitor (named rival / alternative / player), or generic — and flag brand terms being served outside the brand campaign (a "contain_brand" leak). Use this skill whenever the user asks which search terms are branded, which are competitor searches, spots rival brand names appearing in the search terms report, wants to find brand leakage (own brand served by generic campaigns), or needs to apply brand protection negatives. Single-axis brand/competitor classification ONLY — for the full multi-label picture (intent + relevance + cannibalization + brand all at once) use the parent search-term-classifier skill.

Give every Google Ads search term its COMPLETE set of labels at once — brand class (brand / competitor / generic), intent type (informational / comparison / cost / persona / geo / transactional), cannibalization (matched by 2+ keywords + who should own it), and a per-keyword relevance verdict — then roll them up to one recommended disposition per term (add negative / route to owner / add as keyword / hand to content / brand policy / keep). Use this skill whenever the user wants the FULL picture per term, a multi-tag / multi-label classification, "tag/label/classify every search term", "one card per term", to triage the whole search-terms report in a single pass, or says a term is several things at once (branded AND informational AND cannibalized). It COMPOSES the cluster's single-axis skills (relevance, duplicates, opportunity, branded), it does not replace them — a single-axis ask belongs to that sibling; the keep/cut-by-spend verdict belongs to the performance skill.

Build the dollar-quantified "Recommended actions" insights card for the Google Ads Search Terms dashboard — it does NOT classify anything new, it JOINS the per-term tags from the classifier skill with the money from the performance skill (and optional opportunity clusters) and rolls them into one insight row per criterion, each expressed in DOLLARS of monthly optimization potential, leading with a total. Use this skill whenever the user wants the search-terms insights card, the "recommended actions" roll-up, the total optimization potential in dollars, or to turn the tagged search-terms tables into a prioritized money list. Seven lanes: Irrelevant negatives, Wasteful (relevant-but-losing-money) review, Duplicate overlap, Competitor reallocation, Branded recapture, and Opportunity (uncovered intents + new ad angles). It is a SYNTHESIS layer — the judgments stay in classifier / performance / opportunity; this merges their dollars.

For a Google Ads account, judge whether spend is over-concentrated in one keyword match type (Broad/Phrase/Exact) and whether any type is sitting untested at $0 — then flag rebalance vs test-a-type, with each type's cost / conversions / cost-per-conversion side by side. Use this skill whenever the user asks about match types, broad vs phrase vs exact, "am I too heavy on broad", whether to test a match type, match-type efficiency or budget concentration, or audits how keywords are bid — even if they don't say "match type". Judges SPEND CONCENTRATION + per-type efficiency ONLY; per-keyword relevance belongs to `search-terms/relevance` and search-term mining to `search-terms/n-grams`.

N-gram analysis of a Google Ads search-terms report (the Brainlabs search-query-mining method). Breaks every search term into 1/2/3-grams (single words, two- and three-word phrases), adds up the cost / clicks / conversions / value of every term that contains each n-gram, and surfaces (a) WASTE n-grams — words/phrases that spend without paying back → list-level negative-keyword candidates, and (b) WINNING n-grams — strong-converting themes → expand into keywords / ad groups. Use whenever the user wants n-gram or "n gram" analysis, search-term word/phrase mining, bulk/list-level negative keywords, "which words waste my budget", recurring-token waste, theme discovery from search terms, or to find one negative that kills many bad terms. Token/phrase-level aggregate ONLY — single term↔keyword relevance → relevance; per-term keep/cut → performance; cannibalization → term-routing.

Runs on the Porter Metrics MCP (live google-ads data). Read and map a Google Ads account's FULL existing negative-keyword inventory — campaign-level, ad-group-level, and shared negative lists — so every skill that RECOMMENDS a negative (relevance, term-routing, n-grams, brand-incrementality) knows what's already excluded and never suggests redundant work. Use whenever the user asks what's already negatived / "do I already block this" / negative-keyword audit / map existing negatives / "don't recommend things I've already done", or before finalizing any add-negative recommendation. Reads via the connector-action GAQL path (NOT query_data — negatives are config, not metrics). Provides a redundancy check (is a candidate negative already covered, applying Google's exact/phrase/broad negative-match semantics) and surfaces conflicts (a negative blocking a good keyword/term). One account at a time.

Judge the MONEY each Google Ads search term makes or wastes — aggregate every row of the same term, compute cost / conversions / CPA / conv-rate (or ROAS for ecommerce) from the base counts, and classify each term Winning (converts at/under the account benchmark) / Watch (converts above it, or too few clicks to judge) / Waste (real spend, ~0 conversions, past the thin-data floor), ranked by spend, with a destination (promote to exact / leave / add negative) and the dollars at risk. Use this skill whenever the user works the search-terms report and asks which terms are wasting budget, which to promote to exact keywords, which to negativize for poor performance, "what's burning money", cost/CPA/ROAS per search term, or wants the spend-ranked keep/cut read. Performance ONLY — the matching verdict (is this term relevant to the keyword?) is the relevance skill; who should own a duplicated term is term-routing; new keyword/asset ideas are intent-discovery.

Classify each Google Ads placement — the website, app, or YouTube video the ad actually showed on — as on-topic vs off-topic for the advertiser and brand-safe vs unsafe, by reading the placement name / URL and scraping the page when it's ambiguous, then emit the exclusion list and the % of content spend wasted on irrelevant or unsafe content. Use whenever the user asks where their ads are showing, about placements, brand safety, content exclusions, "what sites / videos am I on", wasted Display / Demand-Gen spend, or wants to read the audience from the content it consumes — even if they don't say "placement". Display / Video / Demand-Gen only (Search and Performance Max do not expose placements).

Make a page's images light and SEO-ready in one pass. Converts JPG/PNG to WebP, downscales oversized images, and compresses them (often 70–90% lighter) to speed up load time / LCP — then Claude looks at each image and writes its alt text and title, and wires everything into the HTML (webp src, width/height to stop layout shift, lazy-loading). Use when the user says /image-optimizer, or asks to optimize / compress images, convert images to WebP, make images lighter or the page faster, reduce image weight, or auto-write alt text / image titles / image descriptions for SEO and accessibility. Pairs with /landing-page-audit, which detects heavy or un-described images.

Full technical + performance audit of a landing page (or any URL). Grades Core Web Vitals (LCP / INP / CLS — Lighthouse lab + real-user CrUX field data via Google's PageSpeed Insights API), metadata (title, meta description), indexability (robots.txt, noindex, canonical, sitemap, hreflang, HTTPS, redirects), semantic structure (one H1, heading order, image alt, JSON-LD schema that matches the visible text, Open Graph), and mobile/responsive (viewport, tap targets, responsive images, mobile speed). Every check is verified against the live page and Google's own APIs — nothing is assumed. Use when the user says /landing-page-audit, or asks to audit a landing page's speed / performance / technical SEO, check Core Web Vitals, find why a page is slow, or check whether a page is findable, indexable, mobile-friendly, or technically sound. Works on any public URL, not only landing pages.

Structure and write high-converting landing pages by matching page structure to the traffic's awareness stage and filling each section from the messaging — then hold the copy to a quality bar (components, adjective test, brevity, checklist). Use whenever the user wants to create, plan, write, edit, audit, or critique a landing page, homepage, or product page — its structure, sections, headlines, or copy. Terminal asset of the strategy stack: pulls from positioning-messaging, which pulls from stp-marketing-strategy.