Workflow Breakdown: The PR Pitching Engine

One of the most requested workflows we build is what we call The PR Pitching Engine—an autonomous agent that researches journalists, personalizes pitches, and manages follow-ups without human intervention.

This single workflow has saved our clients an average of 15 hours per week in manual media outreach. That's 780 hours per year—or roughly one full-time employee's worth of productivity.

In this breakdown, I'm going to show you exactly how we build it using modern workflow automation platforms and AI agents.

The Problem: Media Outreach is a Time Sink

Most PR agencies follow the same manual process:

1. Junior hire manually searches for journalists who cover relevant beats
2. Build a spreadsheet of contacts with basic info (name, outlet, email)
3. Account Manager reviews the list and approves targets
4. Junior drafts "personalized" pitches (mostly copy-paste templates)
5. Account Manager edits each pitch for tone and accuracy
6. Send emails and manually track responses in another spreadsheet
7. Follow up 3-5 days later if no response

For a mid-sized PR client with ongoing campaigns, this process consumes 15-20 hours per week. And 90% of that time is spent on tasks a well-designed agent could do better.

The Solution: The PR Pitching Engine

Here's what the agentic version looks like:

The Automated Flow

1. Trigger: New campaign brief submitted via database form
2. Research: Agent searches journalist databases and media monitoring tools
3. Qualify: AI scores each contact against brand guidelines and recent coverage
4. Personalize: AI drafts custom pitches based on journalist's latest articles
5. Review: Top 50 prospects sent to Account Manager for approval (one-click)
6. Send: Emails dispatched via email service with tracking enabled
7. Follow-up: Agent auto-follows up if no response within 4 days
8. Notify: Account Manager alerted via messaging platform only when journalist responds

Total human time required: ~3 minutes per campaign (just the approval step).

Step-by-Step: Building the Engine

Let's break down each component in the workflow.

Step 1: Campaign Brief Trigger

The workflow starts when a new campaign brief is submitted through your database or project management system.

Step 2: Journalist Database Research

The system queries media databases and monitoring tools to find journalists who cover the specified beat. The agent pulls name, outlet, email, social handles, and recent article URLs for context.

Step 3: AI-Powered Scoring

For each journalist, AI analyzes their recent articles and scores them 1-10 based on:

• • Relevance to campaign topic
• • Tone alignment with brand voice
• • Outlet authority and reach
• • Recency of coverage in this beat

Step 4: Filter by Qualification Score

The system filters out anyone with a score below 7. Only the best-fit journalists move to the next step.

Step 5: Personalized Pitch Generation

For each qualified journalist, AI generates a custom pitch that:

• • References their most recent article
• • Explains why this story fits their beat
• • Uses the brand's voice guidelines
• • Includes relevant data points or exclusives

Step 6-8: Approval, Send & Follow-Up

The top 50 pitches are added to an approval queue. When Account Manager marks records as "Approved," the workflow triggers and sends emails through your email service provider. If no response within 4 days, the system checks if the email was opened and sends a follow-up accordingly.

The Results: 15 Hours → 3 Minutes

Here's what one of our clients saw after implementing this workflow:

Why did response rates improve? Because the AI-generated pitches were actually more personalized than the copy-paste templates the junior hire was using. The agent reads every recent article and tailors the pitch accordingly.

Beyond PR: The Same Logic Works Everywhere

This exact workflow can be adapted for:

• → Influencer outreach: Research creators, score by engagement, draft partnership proposals
• → B2B sales: Find decision-makers, personalize cold emails, track follow-ups
• → Podcast booking: Identify shows, pitch guests, manage scheduling
• → Partnership development: Research potential partners, draft co-marketing proposals

The pattern is universal: Research → Qualify → Personalize → Send → Track. If your agency does any variation of this work, it can be automated.

Frequently Asked Questions