Teardown No. 26 · Outbound

How Jasz Joseph went from zero podcast bookings to 10 with a custom GPT her VA runs

Jasz Joseph wanted more podcast guest spots. Her VA struggled to tell which shows fit her technical work. She built a custom GPT that knows her business, decides whether a show is a fit and drafts the email pitch or form answers. It also writes a summary for a shared tracking doc. The story reports 10 bookings in just over two months, up from zero the previous quarter.

Jasz JosephVerified Oct 16, 2026Markdown copy
Part of playbook: Get booked on podcasts with agents, then turn each episode into a content pack

The problem

Jasz set a marketing goal to be a guest on more podcasts. She already had a VA, but the story says “the VA struggled to understand the technical side of what she does to effectively pitch the right podcasts.” For example, Jasz doesn’t do any work with ecommerce or B2C, and the VA couldn’t always spot shows aimed at those audiences.

She describes the judgement involved: “You have to be able to read the podcast description and then really understand: What does Jasz bring to the table that makes sense for this audience?”

The blueprint (as the story describes it)

  1. Feed the GPT the basics. She gave it her website’s homepage, services page and bio. She also listed the types of services she offers and shared examples of past projects. She stated who her ideal clients are and which industries she avoids.
  2. Add the softer side. She told it personal details and values. She used ChatGPT’s voice-to-text feature to talk instead of typing.
  3. Test it yourself before handing it off. Initial setup took “an hour or two.” Testing took about five to eight hours over a couple of weeks. Early pitches weren’t customised enough to each show’s audience, so she trained the GPT to ask the VA a few questions first. The workflow she landed on:
    • The VA researches shows and pastes a podcast’s description into the GPT. The GPT decides whether the show is a fit and explains its reasoning so the VA can override it.
    • If it’s a fit, the GPT asks for the titles and descriptions of the last three to five episodes.
    • The GPT asks whether outreach will go through a web form or an email. For email, it drafts a personalised pitch covering why she’s a fit and what topics she can speak on. For forms, it asks for the questions and writes an answer to each.
    • The GPT writes a summary of the pitch, and they paste it into a shared doc by hand. For forms, the VA takes a full-page screenshot of the submitted answers.
  4. Give feedback like you would to a new hire. She corrected tone and word choice in the GPT itself instead of just editing its output. The VA flags anything that feels off, and they keep refining it in weekly check-ins.
  5. Keep it updated. She adds new client stories, lessons from past appearances and refinements to her speaking topics.

The VA does the final review. He also flags form questions the GPT can’t answer, such as “Who are your three biggest mentors?”, so Jasz can answer them herself.

Research side:

  • She used ChatGPT to compare podcast research tools and chose MatchMaker.fm.
  • She had it help write process docs for the VA. ChatGPT suggested tips she didn’t know, such as using advanced Google search to find podcast leads.
  • She now pastes a batch of podcast links into ChatGPT and asks it to vet them.

Stated results

  • 10 bookings in just over two months since launching the GPT, up from zero the previous quarter. One example is an appearance on Go-to-Market Playmakers.
  • The story says the GPT gave her VA “more autonomy to pitch without additional input” and that pitches now go out without her review.
  • It says the GPT “saved her hours each week.” No hours figure is stated.
  • Setup time: an hour or two to build, plus about five to eight hours of testing over a couple of weeks.
  • No response rate or pitch volume is stated.

Caveats

  • This is a third-party story published on an ActiveCampaign site (autonomous.activecampaign.com), not Jasz’s own write-up. Her quotes come through the article. The text doesn’t promote an ActiveCampaign product. It ends with a “Resources to replicate Jasz’s workflow” heading, and the content under that heading wasn’t checked.
  • The bookings jump is credited to the GPT. In the same period, she also changed her research process with MatchMaker.fm, a paid tool, and bulk vetting. The story doesn’t separate those effects.
  • A human stays in the loop. The VA reviews everything before it goes out, and Jasz answers personal form questions herself.
  • Tracking is manual: the summary is pasted into a doc, and form answers are screenshotted.
  • X: Unverified not checked.

Sources

Want agents like these working your distribution? See the Agent Stack Teardown. Get new teardowns weekly: subscribe (double opt-in).

Get one teardown a week.

Same format, a new operator every issue. Double opt-in.