Teardown No. 43 · SEO

How Deb Szabo went from zero media submissions to weekly ones with a Media Pitching Agent

Deb Szabo wants her name to come back when a business owner asks Google, ChatGPT or Claude who can help build AI systems into their business. She treats third-party media mentions as the evidence AI search reads for authority. Journalist call-outs kept getting buried in her inbox, and deadlines closed before she replied. She built a Media Pitching Agent that scans her inbox twice a day, flags suitable call-outs, prepares submissions, watches for journalist replies and logs everything. The agent never writes her opinion, and nothing is sent without her approval. She reports going from zero submissions to submitting every week. She also reports being quoted in the Sydney Morning Herald and the Daily Telegraph on the same Friday.

Deb SzaboVerified Oct 21, 2026Markdown copy

The problem

Her stated goal for the year was “to be recognised as one of Australia’s top AI business strategists.” She argues that Google’s E-E-A-T framework judges authority by “what independent sources say about you rather than taking your own site’s word for it.” Her own site can show experience and expertise. She lists “four years building AI systems for business, official Claude certifications and over 30 years in marketing.” For authority, she decided she needed media coverage.

Journalists put out call-outs through media request services, and these land in her inbox. She writes: “I was getting these call-outs but they were getting buried in my inbox. By the time I’d go to respond the deadline had closed. It was clear I had a system problem.”

The blueprint (as she describes it)

  1. Write an intelligence file first. The agent uses her “Media DNA,” which she calls “a custom intelligence file,” along with her “AI Brain.” The file covers the subjects she can be put forward for and her point of view on each one. It also holds her proof, results and stories, her credentials and qualifications, and her “boundaries of what I won’t comment on.” She says this is why the agent “can match a call-out to me in seconds.”
  2. Scan the inbox on a schedule. The agent “Scans my inbox twice a day for call-outs that match my experience and expertise in AI, business and marketing.” In her build-your-own list she suggests also monitoring journalist networks, media newsletters and relevant call-out platforms.
  3. Classify each call-out. Her matching rules sort each opportunity as “Strong fit, Possible fit or Skip.”
  4. Alert her and get her input. The agent “Alerts me and flags suitability, gets my commentary and my approval.”
  5. Prepare the submission. It “Prepares the submission including bio, photos, ready to go.”
  6. Keep the opinion human. “What it never does is write my opinion.” She says “nothing goes anywhere without me approving it.” She notes that most media platforms tell users not to send AI-generated responses. She also notes that answering with words that aren’t yours “can get your account blocked.” Her split: “The agent does the admin. Finding, matching, tracking, chasing, logging, formatting.” She does the thinking.
  7. Watch for replies. The agent “Watches for the journalist’s reply so it doesn’t get buried in my inbox.” It alerts her when something needs urgent attention.
  8. Submit and verify. Her build-your-own list says to send the approved response and confirm it was received.
  9. Log and sweep. “Every submission goes into a registry, and a Google Alert tells me the moment something gets published.” She recommends recording submissions, journalist replies, interviews and published coverage.

She says it “handles about ninety per cent of the work, and all of it is admin.”

Stated results

  • “I’ve gone from submitting zero to submitting regularly every week. Newspapers, magazines, podcasts and TV interviews.”
  • She was quoted in the Sydney Morning Herald and the Daily Telegraph on the same Friday: “Two national news outlets on the same day.” The post links to both articles. The Sydney Morning Herald piece does quote her; the Daily Telegraph piece is paywalled and was not checked.
  • She found out about that coverage through the Google Alert step: “That’s how I found out about Friday.”
  • No count of submissions, acceptance rate or AI search recommendation outcome is stated.

Caveats

  • She sells a pre-built Media Pitching Agent inside her paid Content Catalyst program, and the post links to it. The blog also promotes her weekly newsletter.
  • She is an AI business strategist, and this post markets her own services.
  • The two newspaper quotes are her own result claims. The post does not say which submissions led to them or how many submissions she made in total.
  • The post does not name the platform the agent is built on. She refers to “Claude or ChatGPT” generically as the thing that runs the system.
  • She cites an AirOps report on third-party mentions in AI search. Those figures are AirOps’ research, not her results.
  • The “about ninety per cent” figure is her estimate of the share of work the agent handles. It is not measured.
  • Handles (verified 2026-10-21, all linked from debszabo.com): LinkedIn linkedin.com/in/deborah-szabo, Instagram @debszabo_, TikTok @deborahszabo, Facebook facebook.com/debszabomarketing, YouTube @debszabomarketing. Bluesky @debszabo.bsky.social (verified 2026-10-21: name and work match; small account). X: no account linked from her site, and @debszabo and @deborahszabo do not exist on X. Threads, Mastodon: 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.