agentsclimarketplace
Sign in

Builder demand gen

Skill chanktb/claude-google-ads/skills/builder-demand-gen

Builds a Demand Gen campaign blueprint — audience-led, creative-heavy campaigns across YouTube, Discover, and Gmail. Covers audience strategy (your-data, lookalikes, custom segments), image and video creative sets, ad formats, and the upper/mid-funnel measurement caveats. Emits a human .xlsx plus a campaign-spec .json (demand_gen). Reads account-context.yaml. Use when the user says "demand gen", "demand generation", "youtube campaign", "discovery ads", "top of funnel".From its SKILL.md

Install
npx -y skills add chanktb/claude-google-ads --skill builder-demand-gen

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • reads credentialsReads from 1 credential source: `account-context.yaml`.
  • 10 stars10 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.3 KB, 648 tokens by cl100k_base, as published. Nobody here has run it

Google Ads — Builder: Demand Gen

An audience-led, creative-heavy campaign for awareness/consideration. Different from PMax/Search: the audience and creative are the levers, and measurement is upper-funnel.

Operating rules

  • Read context (audiences available, brand, vertical, budget) + the plan output.
  • Write to the working directory: blueprint.xlsx + campaign-spec.json (campaign_type: "demand_gen", asset-group-style creative groups).
  • Set expectations: this is funnel-building, not last-click harvesting.

Model dispatch (run cheap, decide expensive) — see ${CLAUDE_PLUGIN_ROOT}/references/model-tier-dispatch.md

  • Scout (haiku) — per-URL 200 checks; creative-spec/size validation; pulling available audience lists.
  • Routine (sonnet) — pulling customer-match/visitor lists + high-value seeds for lookalikes. Dispatch as general-purpose; return raw lists, don't strategize.
  • Judge (main session) — STEP 2 audience strategy, STEP 3 creative direction (via assets), STEP 4 funnel/measurement framing, bidding. Audience inventory is cheap; audience strategy and creative are judgment.

STEP 1 — Inputs

Funnel goal (awareness/consideration/action), budget, and the products/message to feature.

STEP 2 — Audience strategy

  • Your data: customer-match lists, website/app visitors (retarget).
  • Lookalikes: from high-value customer seeds.
  • Custom segments: by search behavior or competitor/site interest (competitor_terms).
  • Interest/demographic layers as broad context. Combine behavioral + intent; avoid demographics alone.

STEP 3 — Creative sets (via assets)

Build per format: single image, carousel, and video (combined image+video drives more conversions than video-only at similar CPA). Provide briefs/specs and on-screen CTA. Strong, diverse creative IS the targeting lever here.

STEP 4 — Ad formats, funnel role & measurement caveats

  • Map creatives to formats (YouTube in-feed/Shorts, Discover, Gmail).
  • No frequency capping in Demand Gen — monitor reach vs frequency manually.
  • Attribution leans on view-through / brand-lift; don't judge it on last-click ROAS alone. State this.

STEP 5 — Bidding

Maximize Conversions (or value) aligned to the funnel goal; consider brand-lift/awareness objectives for pure top-funnel. Respect ramp discipline + change-event cooldown.

STEP 6 — Output + verification

Emit blueprint.xlsx + campaign-spec.json. Set campaign.status: "paused", conversion_goals.scope: "campaign". Verify creative specs/limits, audience signals present and distinct, Final URLs live. Hand the JSON to pusher (Demand Gen is mostly API-or-UI — guided build in Mode A).

To build / refine later

  • Brand-lift / awareness objective configuration.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most marketing audience skills give in 648 tokens

Counted across 690 of the 894 authors here whose files we hold, read 2026-08-07

  • Apply Poppins font to headingsin 41 of 690, across 6 files
  • Apply Lora font to body textin 41 of 690, across 6 files
  • Use Arial fallback for headingsin 39 of 690, across 4 files
  • Use Georgia fallback for body textin 39 of 690, across 4 files
  • Maintain text hierarchy and formattingin 39 of 690, across 4 files
  • Use accent colors for non-text shapesin 38 of 690, across 3 files
  • Use RGB values for precise color matchingin 38 of 690, across 3 files
  • Use brand colors for primary text and backgroundsin 36 of 690, across 1 file
  • Read product marketing context file before asking questions, starting, or auditingin 35 of 690, across 23 files
  • Use active voice instead of passive voicein 26 of 690, across 10 files
  • Implement or generate appropriate JSON-LD structured datain 24 of 690, across 17 files
  • Prioritize clarity over clevernessin 22 of 690, across 8 files

Said here and by no other author read

  • output both blueprint.xlsx and campaign-spec.json
  • hand the json to pusher
  • read inputs from the account context file
  • set campaign type to demand_gen
  • return raw audience lists without strategizing
  • build single image, carousel, and video creative sets

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.