Surface the objections that block your sales
Your landing page says "streamline your workflow" while your prospects write "I waste two hours a day copying data between tools". That gap costs you conversions. This skill has your AI agent mine Buska's qualified signals and live platform search to find the exact words prospects use for their problem and the objections that stop them from buying, then turn both into copy suggestions backed by linked source posts.
get_signalssearch_mentionsscore_lead1Signals on your category, not just your brand
For this skill, configure Buska signals on your product category and the problems it solves, not only your brand name: objections live in category conversations. Then connect the MCP server (buska.io/mcp) to your agent (setup: buska.io/docs/mcp-server/setup).
2Pull the pain Buska already qualified
The agent calls get_signals with intent set to PAIN, then again with QUESTION, filtered by minScore and a since date. These are posts Buska's engine already detected and scored for your monitored keywords, each with an AI score and the reasoning behind it. Pain signals show how prospects describe the problem in their own words; question signals show what they need answered before they buy.
3Widen the net with live search
Your configured signals only cover your keywords, so the agent also calls search_mentions with your category terms on platforms where buyers talk candidly, like Reddit, G2 or Hacker News. This surfaces raw phrasing and objections around competitors and the category as a whole: comparisons, pricing complaints, trust concerns. Each result comes with content, author, URL and engagement metrics.
4Keep only voices that match your ICP
A loud complainer is not necessarily a buyer. The agent runs score_lead on the most quotable posts to check buying intent and ICP fit, so your copy ends up reflecting the language of people who could actually purchase, not the angriest thread on the internet.
5Your voice-of-customer document
A document in three parts: a vocabulary list of the verbatim phrases prospects use for the problem, the top objections grouped by theme with quotes, and concrete landing page suggestions: headline options, FAQ answers and objection-handling sections. Every claim links back to its source post so you can verify it before anything goes live.
The prompt to paste into your agent
Copy it into Claude, ChatGPT or any MCP client connected to Buska.
Use the Buska MCP tools to help me rewrite my landing page in my prospects' own words.
1. Call get_signals with intent=PAIN, minScore=6, since=2026-07-28, limit=50, then again with intent=QUESTION.
2. Call search_mentions with keyword="cold outreach tool" on reddit (limit=30) and on g2 (limit=30). Swap the keyword for my actual category.
3. From all posts, extract (a) the exact phrases people use to describe the problem and (b) every objection to buying a tool like mine: price, trust, setup effort, switching cost.
4. Run score_lead on the 10 most quotable posts and keep only voices that match my ICP: [describe your ICP here].
5. Deliver a vocabulary list, the top 5 objections with 2 verbatim quotes each (with post URLs), and for each objection one headline option plus an FAQ answer I can put on my landing page.Related documentation
Frequently asked questions
Does this change my landing page or post anything for me?
No. Buska never posts to social networks on your behalf, and this skill produces suggestions, not deployed changes. The agent hands you copy backed by sources; you decide what goes on your page.
Will it work if I have not configured any signals in Buska yet?
Partially. get_signals only returns the qualified signals Buska's engine has detected for your configured keywords, so with none set up that part comes back empty. search_mentions works immediately on any keyword, but you would miss the scored, pre-qualified pain and question posts that make the analysis strong.
Why not just ask an LLM what my buyers' objections are?
Because without data it guesses, and guesses converge on generic objections like "too expensive". Here every phrase and objection comes from a real public post with a URL, so you can read the source, judge the context and quote real language instead of plausible fiction.