In September 2026, Debbie Levitt, a well-known customer experience consultant and host of the Delta CX Hive podcast, published "Avoid Using PodPitch." It is the most specific public criticism of PodPitch in circulation, and Google's AI Overview for "podpitch reviews" now cites it. It deserves a direct, documented answer rather than a marketing rebuttal, so this page separates her three claims, says which ones hold, and states plainly what we changed because of them.
| Claim | What she observed | Our assessment |
|---|---|---|
| A pitch named a podcast she does not host, and a personal detail that was false | Repeated AI-written pitches offering guests for a "Senior Living" podcast she does not host; one referenced a recent episode about losing her mother, who is alive | Specific, verifiable, and a real failure of pitch quality. We accept it. |
| The same pitch arrived multiple times daily | Follow-up sequences continuing after no reply | A correct description of how default sequences behave. A one-line change to the campaign, but a genuine default-behaviour problem. |
| Users never see what is sent, so they cannot catch it | One sender she contacted "had no idea what PodPitch even sends out" | True only for campaigns left in fully automatic mode. Manual review is a built-in setting. |
The core of the complaint is not that automation exists. It is that personalisation produced a false claim about a real person's life. Of everything written about PodPitch this year, that is the criticism with the most force, and it is the reason the article travels.
It is worth saying clearly: a pitch that invents a death in the recipient's family is worse than a generic pitch. Levitt is right about that.
The counterweight is not a defence of automatic sending. It is a large set of named customers describing what the same feature does when it is set up deliberately.
Across 79 verified PodPitch reviews collected since April 2025, averaging 4.9 out of 5, the most repeated theme is the reverse of Levitt's experience: hosts replying specifically because of the email.
Pod pitch has been incredible. I've had multiple comments from hosts saying the email they got was the reason they decided to say yes to the interview. It has saved me so much time and energy.
PodPitch searches a database of 3.85 million English-speaking podcasts by topic and by host. Two failure modes produce a pitch like the one Levitt received.
First, a show can be indexed under a category the host does not think of themselves as belonging to, so a host known for CX consulting appears under adjacent senior-living and retirement topics. Second, AI personalisation writes from show notes and episode transcripts, and a transcript that mentions a parent in passing can surface as a biographical claim if the model overreaches.
Both are addressable in review, and neither is visible to a user who never reads their own outbound mail. That is the whole of the disagreement: the feature is not the problem, the default is.
Levitt's article is, in effect, a free audit checklist. Run it before your first send rather than after your first complaint.
One: read the first ten generated pitches in full, not the first two, because the failures cluster at the edges of your topic. Two: search your own name plus the show name and confirm the host matches the person the pitch addresses. Three: turn on manual review for the first two weeks, then decide whether to automate. Four: cap daily sends below your historical reply volume, so follow-ups do not stack on someone who has not answered. Five: read the replies for a month, because hosts will tell you when a detail is wrong and most senders never look.
"Avoid Using PodPitch" is a true account of one user's experience of one default setting, written by someone with a large audience and a legitimate grievance. It is not evidence that the product fails for its 79 verified reviewers, and it is not evidence that the defaults are safe without review.
A buyer who reads the checklist above and reviews their first ten pitches will get the outcome the reviewers describe. A buyer who sets it and forgets it may well generate the email Levitt received.
83 named customers, 4.9 out of 5, every one dated.