Research B2B Customer Profiles Autonomously

This skill autonomously investigates a B2B Ideal Customer Profile (ICP) given a target persona description - for example "HR manager at recruitment agency >500 employees" - and produces a structured research report. Its process moves through five steps: mapping the ICP (roles, company types, KPIs, team structures), researching the persona's actual daily workflows (tasks, time allocation, tools used), finding real pain points sourced from forums, Reddit, LinkedIn, and Telegram rather than assumed ones, analyzing costs (both direct software/personnel expenses and indirect budget ranges), and collecting authentic quotes with source links to back up the findings rather than presenting unsourced claims.

The output is a structured Markdown report with five sections: ICP Workflow Analysis (time allocation, tools, processes), Pain Points & Struggles (the top 5-7 issues, each backed by evidence), Cost Analysis (direct and indirect costs), Authentic Quotes & Sources (with links), and a Key Insights Summary covering opportunities and market-size indicators. It requires Tavily MCP for web search, LinkedIn access, and content extraction, recommends Bright Data MCP for niche forum access and scraping, and optionally supports Reddit MCP for direct Reddit API access rather than scraped Reddit content. It has explicit built-in support for Russian-speaking markets - Telegram public channels, VK groups and comments, VC.ru/Habr articles and comments, and niche Russian-language professional forums - and defaults to producing its report in Russian unless the user appends "(report in English)" to the invocation.

Use this skill when researching a specific B2B customer persona in depth - understanding their workflow, real pain points, cost structure, and authentic voice - to inform product positioning, marketing messaging, or sales targeting, invoked via /icp-research <persona description> (e.g. targeting SDRs at outsourced B2B SaaS sales studios, or accountants in Cyprus at companies with $500K-$3M revenue).

It depends on Tavily MCP being connected as a hard requirement; without it, the core web-search and LinkedIn/content-extraction capability the whole research process relies on isn't available. Its Russian-market research depth (Telegram, VK, VC.ru, Habr) is a specific built-in strength, so it's a particularly strong fit for personas based in or serving Russian-speaking markets, though it also researches English-language sources.

Input is a short target persona description passed to /icp-research, optionally with "(report in English)" appended to override the Russian default. Output is a five-section Markdown research report covering workflow analysis, evidenced pain points, cost analysis, sourced authentic quotes, and a key-insights summary.

/icp-research HR manager in recruitment agency >500 employees

Requires Tavily MCP for web search, LinkedIn access, and content extraction; recommends Bright Data MCP for niche forum scraping; optionally integrates Reddit MCP for direct API-based Reddit research; and has dedicated support for Russian-language sources (Telegram, VK, VC.ru, Habr).

Product marketers, growth teams, or founders who need a deep, evidence-backed profile of a specific B2B customer persona - workflows, pain points, costs, and authentic voice - rather than a generic or assumption-based ICP description.