How to run podcast SEO customer interviews for audience research

What to take away

Podcast SEO customer interviews are structured conversations with people who already listen to your show or buy from you, run to capture the words they use about the topics you cover.

Recruit from your own list and from people who nearly subscribed, so every answer has context.
Ask about finished episodes and recent searches, not hypothetical wishes.
Group answers into themes, then match each theme to an episode title or page headline.
Keep the script to five questions so a session fits inside half an hour.

Pick the right people and set ground rules

Who belongs in the sample

The criteria below assume an English-language show with an email list and listeners in England. If your audience sits mainly in Scotland, Wales or Northern Ireland, recruit there instead.

Interview sample targets

8-12total interviews to run
2-3near-miss listeners to recruit
0superfans-only samples to avoid

Aim for eight to twelve interviews. Mix regular listeners with two or three people who sampled one episode and stopped, because near-miss listeners expose the gap between your description and their expectations.

Avoid recruiting only superfans. They describe the show in language no newcomer would search for.

To find near-miss listeners, invite people through your newsletter and show notes, then screen for people who sampled an episode and stopped. Copy: “I’m researching how people find podcasts about [topic]. If you tried an episode and didn’t continue, would you be willing to talk about what you expected and what you searched for? I’m offering a £15 voucher for a 30-minute call; reply if interested.”

PodcastFind's Podcast SEO audience research guide for England shows how interviews sit alongside directory data and keyword tools, which keeps a small sample in proportion.

RAJAR keeps UK radio listening data behind a RAJAR subscriber login, though it measures radio rather than podcasts, so treat it as background context.

For wider UK context, Ofcom's Media Nations and Online Nation reports cover media use; they are research reports, not a podcast audience register. Use analytics from the platforms where your show is listed for show-level data.

Consent, incentives and data handling

Recording a call creates personal data, so say what happens to the file afterwards.

The Information Commissioner's Office (ICO) guidance, Direct marketing guidance, covers consent, legitimate interests and opt-out duties if you plan to email participants later. The UK GDPR, the Data Protection Act 2018 and, where electronic marketing messages are involved, the Privacy and Electronic Communications (EC Directive) Regulations 2003 may apply.

For example, a £15 voucher for a 30-minute call is a common thank-you. Keep the amount identical for everyone so payment does not steer answers.

Ask questions that produce searchable language

A five-question core set

Five questions cover the decisions that matter, from discovery to positioning.

“How did you first find this show, and what were you looking for?”
“Which episode did you listen to most recently, and what made you choose it?”
“What words would you type into a search to find another episode like that?”
“Was there anything you expected to find but couldn't, or anything that made you stop listening?”
“How would you describe this show to someone looking for the same thing, and why?”

Five-question core set

How did you first find this show?
Which episode did you last finish?
What would you type to find a show like this?
What nearly stopped you from listening?
How would you describe the show to a colleague?

Ask them in the same order each time, and use neutral prompts rather than offering options.

Capture answers without leading people

Write down exact phrasing rather than a summary. Notes taken during the call beat memory, and a recording made with permission lets you check a phrase later.

Never tidy a quote into marketing copy. The awkward version is the one a listener might type.

PodcastFind's Podcast SEO buyer personas guide explains how to combine these phrases with directory and search data afterwards.

Turn answers into pages and episodes

Code themes into clusters

Group quotes by the job the listener was trying to do, then name each cluster with their own words. A cluster with four or more mentions earns an episode title or a page.

Check search demand before writing. A phrase that lands well in conversation still needs volume and a realistic chance of ranking.

Use Google Search Console to see queries bringing people to your indexed pages, and Google Keyword Planner to check estimated web search demand. Neither measures searches inside podcast apps, so treat them as evidence about web search rather than directory search.

Check that your show is listed in relevant places, including Apple Podcasts, Spotify, YouTube Music, Amazon Music and Podcast Index. For show-level performance, review Apple Podcasts Connect and Spotify for Creators (formerly Spotify for Podcasters); use the listeners' exact phrasing naturally in episode titles and matching pages, and check impressions, clicks and completion rates over time.

Keep cluster names consistent across titles, show notes and landing pages, because repetition helps a search engine connect them.

Before and after changes

Before

Listeners say onboarding, not setup: Episode called setup
Cost questions come first: No pricing page
People search by job role: Described by industry
A rival's phrase is recalled: Competitor wording reused

After

Listeners say onboarding, not setup: Retitled around onboarding
Cost questions come first: Page for the top questions
People search by job role: Notes naming job roles
A rival's phrase is recalled: Own vocabulary used

Each row is a hypothesis, not a proven win. Change one element at a time and watch impressions, clicks and completion rates.

If you promote episodes to a professional audience, LinkedIn's ad tips and best practices page covers targeting and creative formats.

Common questions

How many interviews are enough?

Eight to twelve usually produces repeating themes. Stop when fresh sessions only confirm what you already have.

Should I pay participants?

A small voucher is normal practice and keeps goodwill. Say plainly that payment does not oblige anyone to give positive answers.

Can I quote people in show notes?

Only with permission, and check whether they want their name or employer mentioned. Anonymised phrasing still tells you which words to use.

Does this work for a new podcast?

It works if you have a handful of listeners to approach. With none, interview people in your target audience about the problem your show addresses.