Audience Architect
From target audience to buyer personas, messaging and ad targeting — all in one skill.
SKILL.md in the open agentskills.io standard — works directly in Claude, ChatGPT/Codex, Cursor, Copilot & more.
Audience Architect is an audience-research AI skill that builds buyer personas and an ideal customer profile (ICP) from your product: sharp customer segments and persona cards (jobs-to-be-done, buyer journey, negative list, confidence) plus voice-of-customer analysis, ad targeting and per-segment positioning. B2B/B2C preset included.
What this skill does
- B2B/B2C/B2B2C preset + market/language profile: adapts framing, channels and wording
- ICP & persona cards by established frameworks (STP, JTBD, 5 Rings, optional Empathy Map / VPC)
- Visible prioritization score per segment (value × reachability × fit) as a table
- Buyer journey per persona (awareness → consideration → decision) with a message per stage
- Negative list / anti-persona as its own block (saves ad budget and support)
- Living personas: update existing analyses with new data via carry-forward
- Confidence per persona + [evidenced]/[assumption] split — no invented numbers or profiles
- Voice of customer: reviews/tickets/surveys → pains, real customer language, sentiment
- Ad targeting: persona → concrete LinkedIn/Meta/Google parameters + creative angles
- Messaging & positioning per segment; research design for unbiased surveys/interviews
- Persona cards exportable (table / Notion / Empathy Map / Value Proposition Canvas)
Description
Audience Architect is a skill for target audience research and customer segmentation — it turns your product into sharp segments, buyer personas, messaging and ad targeting, from jobs-to-be-done to positioning. A complete audience toolkit in one skill, cleanly separated by a mode router, with a B2B/B2C/B2B2C preset and a market/language profile. ICP/persona mode built on established frameworks (STP, jobs-to-be-done as a statement, 5 Rings of Buying Insight, optional Empathy Map and Value Proposition Canvas): 2–4 prioritized segments with a visible score table (value × reachability × fit) and persona cards (JTBD, pains/gains, triggers, objections, decision criteria, channels, buyer journey, hooks), including buying-committee roles for B2B. A dedicated negative list / anti-persona saves ad budget and support. Living personas: existing analyses are updated with new data (carry-forward). Voice of customer mines reviews/tickets/surveys for pains, real customer language and sentiment. Targeting translates a persona into concrete LinkedIn/Meta/Google parameters plus creative angles. Messaging/positioning delivers value props, hooks and positioning statements per segment. Research design creates unbiased surveys and interview guides. Honest throughout: separates evidenced from assumed (with a confidence level), invents no market sizes/statistics/quotes, builds no fictional profiles, avoids stereotypes and works GDPR-aware. Persona cards are exportable (table, Notion block, Empathy Map, VPC).
Examples
What it does not do
- Invents no market sizes, demographic shares, statistics or customer quotes
- No fictional persona profiles (made-up age/photo/biography) — evidence-based, not invented
- No stereotypes or discriminatory generalizations; GDPR-aware (no PII spreading)
- Won't pass off simulated persona reactions as real research
- Won't run ad accounts, won't write the article, won't parse raw tables (use other skills for that)
Compatibility & tech
- Tested (internal)
- 3 scenarios
- Recommended runtime
- Strongest available model; with a CSV/web tool for data-driven analysis.
- Modes
- Profile/ICP · Voice of customer · Targeting · Messaging/positioning · Research design
- Inputs
- Product description · Customer data (CSV) · Reviews/tickets · Survey answers · Existing persona
- Output format
- Persona cards (evidenced/assumption + confidence, with buyer journey) + prioritization table + negative list; depending on mode VoC, targeting, messaging or research design; exportable
- Subcategory
- Audiences & positioning
- License
- Proprietär
Security profile
Runs entirely on your machine with your own AI — no external runtime, no running costs.
Only instructions, templates and references — no executable scripts.
Works offline with what you provide — does not call external services on its own.
What you get
- audience-architect-1.4.0/3 files
- .claude-plugin/marketplace.json
plugins/audience-architect/skills/audience-architect/3 files
- SKILL.md
- manifest.json
references/1 file
- audience-architect-checklist.md
- .agents/skills/audience-architect/→ universal — same content (Codex, Cursor, Copilot, Gemini, Windsurf, Cline)
- LICENSE.txt
Installation
Also works as a chat prompt
No AI tool? Paste it into Claude, ChatGPT or Gemini and use the method right away.
This skill ships no scripts, so the prompt carries its full method.
Installing is the full version — it triggers automatically, runs its scripts and loads references as needed. As a chat prompt you drive the method by hand.
Unlock to copy the ready-to-paste prompt — then in “My Skills”.
Reviews
No reviews yet – be the first.
Note
Strongest with real data (customer list, reviews, survey) or a web tool for market context; without data the skill works from clearly flagged assumptions with a confidence level.
Changelog
- v1.4.022.06.2026Platform-native output: now actively uses your platform's native formatting — tables, checklists, code blocks and clickable/button options, with a numbered fallback on non-interactive platforms. Plus internal metadata that prepares the upcoming agent builder.
- v1.3.018.06.2026Reworked to the AUTHORING v2 standard: realistic input/answer examples (incl. a calibration case), copy checklist(s) and internal evals; cross-references added.
- v1.2.010.06.2026Buyer journey per persona; visible prioritization score (value × reachability × fit) as a table; living personas (carry-forward); market/language profile.
- v1.1.010.06.2026B2B/B2C/B2B2C preset + buying committee; negative list/anti-persona prominent; established frameworks (STP, JTBD, 5 Rings, Empathy Map, VPC); persona cards with confidence + export; rule against fictional profiles.
- v1.0.010.06.2026Multi-mode audience skill: ICP/personas, voice of customer, ad targeting, positioning/messaging, research design; evidenced-vs-assumption discipline, GDPR- & bias-aware.
Frequently asked questions
What does Audience Architect do?
Audience Architect is an audience-research AI skill that builds buyer personas and an ideal customer profile (ICP) from your product: sharp customer segments and persona cards (jobs-to-be-done, buyer journey, negative list, confidence) plus voice-of-customer analysis, ad targeting and per-segment positioning. B2B/B2C preset included.
How do I create a buyer persona?
A solid persona covers jobs-to-be-done, the buyer journey and a do-not-target list, not just demographics. Audience Architect builds such persona cards including a confidence rating from your product and optional real data.
How do I define my target audience?
Audience Architect turns your product into sharp segments and persona cards and adds positioning and targeting per segment — from the first audience question to a finished analysis.
What is a voice-of-customer analysis?
Voice of customer examines how customers describe their problem in their own words. Audience Architect includes this analysis so messaging connects to real customer language.
How does audience analysis differ for B2B and B2C?
B2B and B2C differ clearly in decision-makers, journey and messaging. Audience Architect ships with a B2B/B2C preset and adapts segments and targeting accordingly.
Which AI tools does Audience Architect work with?
Claude (Projects & system prompt) · ChatGPT (Custom GPT & custom instructions) · Any LLM · Optional: CSV input (customer data) + web tool (market context)
How do I use Audience Architect?
Audience Architect is a SKILL.md in the open agentskills.io standard: install it with one command (npx) or download it and add it to your AI tool — Claude (Projects), ChatGPT (Custom GPT), Cursor, Copilot, Gemini CLI and more. No code needed.