A CEO just sent 2,000 AI-generated podcast pitches in a single batch. He got 112 positive responses. His agent scheduled 35 recordings. He came back from the weekend and his calendar was solid, 10 hours a day, back to back, for an entire week.
This isn't hypothetical. It already happened.
Cody Schneider, CEO of Graphed, went on the Open Market podcast and laid out exactly how he did it. He scraped a podcast database for contact info, validated the emails, loaded them into a cold email tool, and connected an AI agent to manage the replies and schedule recordings.
If you're in the tech or startup space on X, you probably already know Cody. He's built a large following by sharing aggressive, automation-first go-to-market playbooks, and he's not shy about the controversial parts. He openly told the hosts that the AI outreach that booked him on their very show was the same system he was describing.
It worked. It also went sideways in a few entertaining ways: the agent responded to hosts within 12 seconds (immediate bot detection), it agreed to purchase a host's books, and it dropped the ball on 90 of the 112 positive responses. Hosts were, in Cody's words, "viscerally mad."
But here's the part that should matter to anyone in podcast PR: Cody doesn't see any of that as a reason to stop. He sees it as version one. The mistakes get fixed in code and redeployed within 24 hours. The agent gets better. And the next batch will be bigger.
He's already thinking beyond podcast outreach. He described a vision where an AI agent maintains ongoing email relationships with 100,000 people, just talking, building rapport, gathering information, with no immediate intent to sell. And then, when the moment is right, nudging the conversation.
This is where podcast outreach is headed. And if you're a PR professional, you need to understand what it means for you.
This post covers:
- What actually happened when someone automated podcast outreach at scale
- Why the coming wave of AI pitches will change how hosts filter their inbox
- Where AI genuinely helps PR professionals do better work
- Where it makes you worse
- Why relationships will be all that matters, and the numbers that prove it
What happened when someone actually did this
Let's be specific about what Cody built, because this isn't a thought experiment.
His normal process: go to Rephonic (a podcast database), find shows in his category, scrape the contact info, validate emails with Million Verifier, load them into Instantly (a cold email tool), pitch himself, and manage the replies.
The AI version: same process, but an agent handles the reply management and scheduling. He sent 2,000 pitches. The agent tried to book him based on positive responses. It scheduled 35 recordings.
The problems were real. The agent replied too fast, hosts immediately knew it was a bot. It couldn't handle nuanced conversations. It got manipulated by a host into agreeing to buy books. Some hosts were furious.
But Cody's takeaway wasn't "this doesn't work." His takeaway was "this is version one, and it's already working well enough to fill my calendar for a week."
He's not wrong about the technology. The iteration speed is real. The costs are dropping. And he told the Open Market hosts he's already building the next version, agents that run in perpetuity, scraping, pitching, following up, and scheduling without any human in the loop.
The question isn't whether this approach works. It's what happens when thousands of people start doing it.
The flood is coming
Cody is a technical CEO who built his own agent stack. Most people can't do that today.
But the tools are getting easier. Fast. Cody himself said it: the cost of automating a workflow like this has dropped from tens of thousands of dollars to basically the price of API access. He's sharing the exact playbook to a large audience. Other founders and growth marketers are paying attention.
Now scale that across an entire ecosystem of founders, marketers, and agencies who hear this and think: I should be doing this too.
Podcast hosts are about to experience what email inboxes experienced when cold email tools went mainstream. The volume explodes. The quality drops. And the hosts who used to read every pitch start building walls.
Filters. Auto-delete rules. Intermediaries whose job is to screen out AI-generated outreach. Or the simplest solution: stop accepting cold pitches entirely and only book guests through personal referrals.
The irony is vicious. The easier it gets to send pitches, the harder it gets to land them.
Why this actually favors PR professionals
Here's what the "automate everything" crowd misses: podcast outreach has never been a volume game.
Cody sent 2,000 pitches and got 112 positive responses. That's a 5.6% positive response rate. For cold email, that's decent. But let's look at it from the host's side.
If Cody's batch hit 2,000 shows, those hosts each received at least one AI-generated pitch from him. Now imagine 50 founders running the same playbook in the same category. That's 50 AI pitches per host, per cycle, on top of whatever organic pitches they're already getting.
Hosts won't scale up their reading. They'll scale down their tolerance.
And when that happens, what gets through?
The pitch from someone the host recognizes. The pitch that references a specific recent episode, not the show description, but something that aired last week. The pitch from a professional email with a real track record. The follow-up that arrives at the right time with the right context.
In other words: the stuff that PR professionals already do well.
When every pitch sounds the same because it was generated by the same underlying model, the pitch that clearly wasn't is the one that stands out.
The numbers prove it
We see this play out in real data at Podseeker every day.
PR professionals on our platform who do the work, researching shows carefully, matching the right client to the right podcast, pitching with a specific angle, consistently see open rates above 80% and reply rates above 60%. Some users sustain reply rates above 80% across hundreds of emails over months of outreach.
These aren't people sending 2,000 pitches in a batch. They're sending dozens of carefully targeted pitches per campaign, to shows they've actually researched, with angles tailored to each host's audience.
The difference in outcomes is massive. High-volume automated outreach gets a 5% positive response rate and angry hosts. Targeted, relationship-driven outreach gets 60-80% reply rates and ongoing conversations that turn into bookings across multiple clients over time.
This isn't a philosophical argument. It's math. And the math only gets more lopsided as AI-generated pitch volume increases and hosts raise their filters.
Where AI actually helps you
I'm not anti-AI. We use AI extensively at Podseeker. PR professionals who ignore these tools will fall behind.
But the key is knowing where AI creates leverage versus where it creates noise.
Research
This is the biggest win. AI can help you understand a podcast's typical guest profile, recent topics, audience size, and format, work that used to take 15-30 minutes per show. It can surface podcasts you wouldn't have found manually. A podcast database built for outreach uses AI behind the scenes to keep contact data current, generate booking intelligence, and match clients to relevant shows, so you spend your time on judgment, not data entry.
Follow-up tracking
Knowing when to follow up, with whom, and what you said last time. This is workflow, and AI-powered tools handle it well. No pitch should ever fall through the cracks because you forgot.
Relationship memory
You pitched a host six months ago and they said "not now, try me in Q1." Your system should remember that. AI makes this effortless.
Where AI makes you worse
The pitch decision itself
The moment you automate away the actual pitch, the decision of "is this guest right for this show, and here's why," you become indistinguishable from spam.
Cody's story proves this. His agent couldn't tell when a host was a bad fit. It couldn't handle a host who pushed back. It couldn't even avoid getting talked into buying books. The sending worked. The judgment didn't.
And hosts are learning to spot AI pitches fast. They see the same structures, the same flattery openers, the same generic angles. When your pitch reads like it was fed the show's Apple Podcasts description and nothing else, that's not personalization. That's a mail merge with extra steps.
Bulk sending and deliverability
Email deliverability is reputation-based. Cody used Instantly, a cold email platform designed for high-volume outreach. It works, until your domain's sender reputation tanks from hosts marking your pitches as spam.
If you're a PR professional, your email domain is your business. A damaged sender reputation doesn't just affect one campaign. It affects every client you pitch for.
Removing yourself from the relationship
This is the one that matters most long-term.
Podcast outreach is relationship work. A host who books your client today might book your next client in three months, if you've built trust. If your first interaction was a pitch they could tell was automated, that relationship never starts.
Cody acknowledged this tension on the pod. He framed it as a tradeoff between efficiency and brand risk. But for PR professionals, your brand is the whole business. You can't afford for hosts to associate your name with bot outreach.
The playbook that wins in 2026
If you're a PR professional or podcast booking agent, here's how I'd think about this.
Use AI to compress your research phase. What used to take an afternoon should take an hour. Find podcasts faster. Understand them deeper. Surface the non-obvious matches that your competitors miss.
Keep humans in the pitch decision. The act of deciding "yes, this podcast is right for this client, and here's the specific angle" stays human. That's your value. That's what hosts are filtering for.
Build systems, not blasts. The goal isn't to send more pitches. It's to never lose track of a conversation, never miss a follow-up window, and build a reputation with hosts over time so they recognize your name in their inbox. Workflow tools matter far more than generation tools.
Pitch from your real identity. Professional email. Real name. Track record of booking relevant guests. In a world of anonymous AI-generated outreach, your identity becomes your moat. Hosts will increasingly whitelist known, trusted PR contacts and block everything else.
Follow up like a human. Thoughtful. Spaced. Referencing something specific. Not a sequence that fires automatically regardless of context.
Think in placements, not pitches. One strategic podcast placement on the right show at the right moment can generate more value than a thousand automated emails. That's always been true. It's about to become obvious.
How Podseeker fits into this
We built Podseeker around a specific belief: the best outreach is human judgment amplified by good tools, not replaced by automation.
That's why we built it the way we did:
AI helps you research and draft. When you find a podcast in Podseeker, you can generate a personalized pitch draft using your client profile and the show's recent content. But you review and approve every pitch before it sends. Nothing goes out automatically.
The workflow keeps you organized. Every pitch has a next step: follow up, reply, snooze, archive. Your pitch workflow tells you exactly what needs attention. No spreadsheet. No guessing.
Follow-ups are smart, not robotic. Schedule follow-ups individually or across an entire campaign. Podseeker cancels them automatically if the host replies first, so you never send an awkward "just checking in" to someone who already said yes.
Contact data is accurate. Our internal AI agents scrape podcast websites directly and keep contact information current. You pitch the right inbox, not a dead email from three years ago.
The philosophy is simple: use AI where it saves you time. Keep yourself in the loop where it matters. For a full walkthrough of how this works in practice, read: How to Use Podseeker.
Relationships will be all that matters
Cody said something on the pod that stuck with me: "Having good ideas is the hardest part. The middle work can be solved."
He's right, but only partially. In podcast outreach, the "middle work" is the sending, tracking, and scheduling. That absolutely can and should be automated.
But the relationship, the trust you build with a host over time, the reputation that makes them open your email instead of deleting it, the judgment to match the right guest to the right show, that's not middle work. That's the work. And no agent can do it.
We're heading toward a world where AI handles all the logistics of outreach and the only differentiator left is whether a host wants to hear from you. That's a relationship question, not a technology question.
The PR professionals who'll thrive aren't the ones racing to send 2,000 pitches in a batch. They're the ones who use better tools to research faster, track everything, never miss a follow-up, and then apply their own judgment to the part that actually matters.
When every inbox is flooded with AI-generated noise, the signal has never been more valuable.
Ready to see what human-led, AI-assisted podcast outreach looks like?
Start Your Free Podseeker Trial →
Try us risk free with a FREE 3 days trial.
