How to Personalise a Podcast Pitch at Scale

Personalisation and volume pull against each other, and no tool resolves that. What scales in podcast outreach is everything except the personalisation: the angle, the proof, the structure, the ask, the follow-up timing. Those get built once. The two or three sentences that establish fit with a specific programme still cost ten to twenty minutes of listening, and there is no version of this where they do not.

So the practical question is not how to automate that part. It is how to make sure those minutes go to shows where they change the outcome.

Separate the fixed parts from the variable parts

Open your best-performing pitch and mark every sentence as fixed or variable.

Fixed sentences are true regardless of who receives them: your angle, the experience that supports it, what a listener would take away, your ask, your sign-off. These are written once, edited a few times over a couple of months, and then left alone.

Variable sentences depend on the show: why this programme, why now, and how your angle connects to something the host has already covered or deliberately avoided.

In a working pitch the split is roughly 80 percent fixed and 20 percent variable. That ratio is the reason this can be done at any volume at all. It also shows where the leverage sits, because improving a fixed sentence improves every pitch you send from that point forward, while improving a variable sentence improves one.

Spend a full afternoon on the fixed 80 percent. Most people spend that afternoon on the variable part of a single email and never revisit the sentences going out fifty times.

Tier the target list

Not every show deserves twenty minutes. Sort the target list into three tiers before you write anything.

The tiering does two useful things. It stops you spending A-tier effort on a show you would decline if they said yes, and it stops you sending a C-tier email to the show you actually want.

Be honest about tier C. Those pitches are close to generic, and the reply rate reflects that. They are worth sending only when the marginal cost is genuinely low and you have already covered A and B. Sending forty C-tier emails is not a substitute for fifteen A-tier ones.

Build a research routine you can run in ten minutes

The research is repeatable even though the output is not. Run the same five steps for every show and the time collapses.

1. Read the last twenty episode titles. This tells you the recurring themes and whether your angle has already been covered.
2. Read two full episode descriptions. This tells you how the host frames a conversation and how technical they go.
3. Check the guest or contact page. This tells you whether they take pitches and how they want to receive them.
4. Note one specific thing: a question the host keeps returning to, a claim you can complicate, a gap in the coverage.
5. Record the contact route and the date.

Step four is the only one producing the variable sentences, and it is the only one that cannot be delegated to a tool. The other four are lookup, which is exactly where a well-built directory saves time. Our podcast directory exists to compress steps one, three and five so the listening time survives.

Reuse angles, reframe the tension

Carrying one angle per show does not scale. Carrying twenty does not work either, because you lose track of which belongs where.

Two or three angles is the working number. Each one attaches to a different part of your experience, and each gets reframed rather than rewritten for the show receiving it.

Reframing means keeping the underlying claim and changing the tension it is stated against. The same experience of migrating a pricing model reads as a finance story for one audience, an operations story for another and a customer trust story for a third. Identical facts underneath, a different sentence in front of the host.

That is genuinely reusable, and it is different from a merge field, because the reframe requires you to know what that audience worries about. Building the angles themselves is the upstream work that makes this possible.

What to automate and what to leave alone

Automate the record: show, date contacted, tier, angle used, follow-up due, outcome. A spreadsheet handles this. Missing follow-ups is a far more common failure than weak personalisation, and it is entirely a tracking problem.

Automate the drafting starting point. Getting to a first draft you then cut down is faster than facing an empty message, which is what the pitch generator is for.

Do not automate the send. Campaign tooling adds tracking pixels and unsubscribe headers that identify the email as list traffic, and that identification undoes the personalisation you just paid for. Send from your own mailbox.

Do not automate the variable sentences. Text generated about a show nobody listened to produces confident, plausible, slightly wrong claims about that programme, and being wrong about a host's own show is worse than saying nothing specific at all.

The realistic ceiling

For a founder doing this alongside a real job, ten to twenty properly researched shows a week is the sustainable rate. That is 40 to 80 a month, which is enough to build a genuine baseline and enough to produce bookings if the angles are sound.

The way to go beyond that is more people doing the research, not a better template. That is the actual reason booking agencies exist, and it is worth saying plainly rather than pretending a system fixes it.

Where to take this next

Spend one afternoon on the fixed 80 percent of your pitch. Then tier your next forty shows and put the research minutes into tier A.

If the honest answer is that ten hours a month is not available, Convokast does the research, the target list, the pitching and the follow-up for $499 a month, with at least one guaranteed interview a month and your approval on every show before it is pitched.