Get an API key
Sign in to Klinko, then generate and securely store a test or production API key.
Audience discovery skill
Audience FinderFind who is most worth serving first.
Find and rank audiences with Klinko Audience Finder, powered by the Klinko MCP integration for Codex and Claude Code.
KLINKO / SKILL MAPAudience discovery
Direct answer
What is Audience Finder?
Audience Finder is an AI market research skill that organizes current market evidence into a ranked shortlist of candidate audiences. It separates visible groups from commercially relevant ones, explains who is most worth serving first, and shows what evidence or validation could change that decision.
Use it in Codex or Claude Code
Connect once. Ask in natural language.
Klinko Skills run through the authenticated Klinko MCP. Add your API key once, connect the MCP client, and use this Skill without learning a command language.
Connect Klinko MCP
Follow the GitHub installation guide to add Klinko MCP to Codex or Claude Code.
Run Audience Finder
Describe the market question directly; the Skill applies a consistent research standard and output.
ENV
KLINKO_MCP_API_KEYThe Free plan includes a one-time balance of 200 credits. View pricing. Never place the API key in prompts, logs, or a public repository.
What this skill helps you decide
Find who is most worth serving first.
Use it when “our customer could be anyone” is blocking product, positioning, or acquisition decisions. The skill creates a defensible starting point without pretending market evidence is a representative census.
Inside the skill
Audience Finder research process
Inputs
- Product, offer, or problem statement
- Market and geographic scope
- Known customer clues or exclusions
- Decision constraints such as budget or time
Decision process
- 01
Identify candidate groups
Map recurring identities, situations, jobs, and problem language in the supplied market scope.
- 02
Separate visibility from value
Distinguish frequent discussion from urgency, buying motivation, fit, and ability to act.
- 03
Prioritize the shortlist
Rank candidate audiences and state assumptions, contradictory evidence, and the next validation.
Outputs
- Prioritized audience shortlist
- Segment evidence and confidence notes
- Needs, triggers, barriers, and language
- Recommended next validation step
Natural-language guide
Start with one market question
Use the complete question, then ask for evidence boundaries and the next test. The Skill keeps the research path consistent while the market question changes.
We are building an AI meeting assistant for small professional-service firms. Which audiences should we investigate first in the US?
Separate conclusions supported by public evidence from inferences, and show what information is still missing.
What new evidence is most likely to change the recommendation, and what should I validate next?
Worked example
Market question
QUESTIONWe are building an AI meeting assistant for small professional-service firms. Which audiences should we investigate first in the US?
What the skill clarifies
The skill may separate client-facing boutique consultancies, specialist agencies, and small legal teams instead of treating “small business” as one audience, then rank them against observable urgency and workflow fit.
Recommended actions- Interview the top-ranked segment around the riskiest assumption
- Test its problem language on one landing page
- Keep the runner-up as a deliberate comparison group
- You have a product idea but no credible first audience
- Several plausible customer groups compete for attention
- Your current persona is broad, inherited, or assumption-led
- It prioritizes research targets; it does not prove market size or product-market fit.
- market evidence can underrepresent private, offline, or low-posting populations.
FAQ
Questions about this skill
What does Audience Finder produce?
A ranked audience shortlist with supporting signals, motivations, objections, assumptions, and a next validation step.
Is it a buyer persona generator?
No. It first decides which audience deserves investigation. A persona should be built only after that choice has evidence.
Can it replace customer interviews?
No. It narrows whom to interview and what uncertainty the interview should test.
Can I use Audience Finder with Codex and Claude Code?
Yes. Audience Finder supports Codex CLI, the Codex desktop app, and Claude Code, using Klinko MCP for market evidence.
Continue your research
Use the right skill for the market decision in front of you.
Connect Klinko to Codex or Claude Code, then start with a focused market question.