Signal Documentation

Signal refreshes its journalist and publication data every 48 hours. This means every journalist's recent articles, beats, and coverage patterns are re-analyzed twice a week. Unlike databases that update monthly or quarterly, Signal's data reflects what journalists are covering right now.

Fit Score is Signal's proprietary match quality metric. It uses 1024-dimensional embedding vectors (via Amazon Titan Embed v2) to measure the cosine similarity between your expertise vectors and a journalist's recent coverage.

A Fit Score of 90%+ means the journalist has recently covered topics deeply aligned with your expertise. Signal recommends pitching journalists with a Fit Score above 85%.

The Knowledge Base stores your company's expertise context — press kits, product descriptions, key metrics, and talking points. This context powers AI pitch generation: when Signal drafts a pitch, it references your Knowledge Base to write pitches that sound like you, not generic templates.

Architecture

The MCP bridge is a lightweight Node.js proxy that translates MCP tool calls into Signal Core API requests:

Navigate to Settings → Knowledge Base and paste your company URL. Signal's AI crawls your website and extracts 5 expertise vectors.

Open Pitch Radar from the sidebar. Signal has already scanned journalist requests and ranked them by Fit Score against your expertise vectors.

Click on any opportunity to see the journalist's recent coverage, their beat, and an auto-generated pitch draft. The Fit Score tells you how well-aligned this journalist is with your expertise — anything above 85% is a strong match.

Every opportunity comes with a pre-generated pitch draft. The AI uses your Knowledge Base context + the journalist's recent coverage to write something specific — not a template.

Edit the pitch inline. Signal writes in "Expert Industrial" voice by default: short, data-backed, zero fluff. Adjust the tone and add your own angle before deploying.

# Example auto-generated pitch:

"Given your deep-dive into AI cost optimization last week, 
you'll want to know that our platform just reduced inference 
costs by 40% for 3 Fortune 500 clients — happy to share the 
methodology exclusively before the Q3 earnings cycle."

# No templates. Lead with data. Under 200 words.

Click Deploy Protocol to open the deployment wizard:

1. Select Muscle — Pick your connected outreach tool (Instantly, Smartlead, or Apollo)
2. Select Campaign — Choose which campaign to add the lead to
3. Preview — Verify the pitch variables and toggle Draft vs. Active mode
4. Dedup Check — Signal checks if you've pitched this journalist in the last 30 days
5. Push — The lead is added to your outreach campaign

Signal shows a queue status indicator for all deployed pitches. You can see which leads were pushed successfully, which had duplicates, and which failed — all without leaving the Signal dashboard.

As new journalist requests come in (refreshed every 48 hours), Pitch Radar automatically surfaces new opportunities. Set up Topic Trackers in Settings to get morning briefs with the most relevant matches.