A journalist request can close within hours. The useful result is not a large media list. It is a small queue of current requests that match expertise your company can support with facts.
SocialListeningAPI searches public X and LinkedIn posts plus Google results through the MCP. It returns normalized records with source URLs. An external workflow handles scheduling and delivery, while a person verifies the journalist and deadline.
Step 1: Define the approved expertise
List the subjects where someone can give a useful, attributable answer. Add the approved expert, role, evidence, and topics they should not discuss.
Build searches such as:
looking for sources industryseeking expert comment topicjournalist request industryneed a source topicdeadline expert comment
Keep the request wording separate from the industry term so you can see what found each result.
Step 2: Search media requests through the MCP
Create a SocialListeningAPI account, claim the 100 free credits, and connect the MCP to Claude or ChatGPT through OAuth.
Ask Claude or ChatGPT to preserve the source details:
Search X, LinkedIn, and Google for public journalist or expert-source requests about the approved
topics. Return the requester, outlet when stated, topic, stated deadline, requested expertise,
published date, and source URL. Leave missing fields blank. Do not contact anyone and do not invent
an outlet or deadline.One query across these three sources costs up to 4 credits: 2 for LinkedIn and 1 each for X and Google.
Step 3: Route new requests externally
Use an external MCP-capable agent and scheduler for repeated searches. Store source URL, requester, topic, published date, stated deadline, first-seen time, and review state. Skip duplicate URLs.
Send new matches to email, Slack, a sheet, or a PR system outside SocialListeningAPI. The MCP does not provide native scheduling, alerts, identity verification, or outreach.
Step 4: Verify and answer as a person
Open the source and verify the requester’s identity, outlet, topic, and deadline. Confirm that the expert has direct knowledge and can support the answer. Skip vague requests, expired deadlines, and subjects outside the approved list.
Keep the pitch short. Answer the request, state the expert’s relevant role, and offer evidence. Do not send a generic company introduction to every result.
Conclusion
The expected output is a small queue with topic, deadline, requester, and source URL. Speed matters, but a verified fit matters more than replying to every request.
Continue with Build Brand Mention Alerts With an API.
