Podcast SEO audience research: surveys and interviews

What to take away

For podcast SEO, match the method to the questionsurveys count how common a behaviour is; interviews explain why it happens.
Anchor the work in public evidence, including RAJAR listening volumes, ONS leisure and tourism releases and IAB UK mobile guidance.
Convert each listener need into one search query, one episode title and one opening line of show notes.
Score methods on decision relevance, cost, speed, evidence quality, repeatability and compliance effort.
Listener data is personal data under the Data Protection Act 2018, so plan consent, storage and deletion first.

Read the public evidence before you commission anything

RAJAR publishes quarterly listening volumes showing how much audio UK audiences consume and on which devices. Use it to size the audio habit, not your niche.

ONS leisure and tourism statistics cover how households spend free time and money. That frames podcast listening against streaming, gym memberships and nights out.

IAB UK reported that brands need to fit their mobile content into women's daily routines. Assume a glance on a phone screen, not a careful read.

Decide between a survey and an interview

Write the research question before choosing a method. A question beginning "how many" points to a survey. One beginning "why" or "which words" points to interviews.

Surveys give proportions you can quote to a sponsor. Interviews give the exact phrasing listeners type into a search box.

Sample fit beats sample size. Twenty respondents drawn from your own listening audience tell you more than two hundred strangers.

Before choosing any survey platform, decide what result would change an episode title or content plan. This keeps the research tied to a search decision rather than collecting data for its own sake.

Survey or interview?

Need a number or share? Yes -> run a survey

Yes

Need exact listener language? Yes -> run interviews

Need both? Run interviews first, then survey

Example: from a listener need to a keyword and a title

Illustrative example, not a research finding.

Suppose a business podcast keeps hearing that subscribers open the newsletter but never book a call. That is a need. It is not yet a keyword.

The query a listener types sits closer to "why newsletter subscribers do not book a call" than to "email conversion optimisation".

Episode title: "Why newsletter readers do not book: fixing the ask, not the audience."

Show-note opening:

If your newsletter gets opened and your calendar stays empty, the problem is usually the ask, not the audience.

Need to keyword to episode

List the need
Write the listener's question
Turn it into a candidate keyword
Assign one theme per episode
Revisit the mapping quarterly

Carry that phrasing into the first 100 characters of the description and the episode page title tag. Search engines and podcast directories both read it.

Score methods before you spend

Criterion	A score of 5 means	Weight
Decision relevance	Findings change a title or content plan this quarter	3
Evidence quality	Data comes from your own listeners or a named public source	3
Cost	Under £200 or under five hours of team time	2
Speed	Results inside two weeks	2
Repeatability	The same method can run next year for comparison	2
Compliance effort	Consent, storage and retention stay simple to document	1

Multiply each score by its weight and add the column. Set the weights before you score, and use the rubric at planning stage, not afterwards.

Survey questions and an interview script you can reuse

After each interview, record the listener's need, exact words, app or search context, and what prompted them to press play. Remove identifying details from working notes, then group similar needs while retaining the original phrases. Compare each theme with survey search terms; turn a theme into a candidate query, then choose a title that answers it and open the show notes with the problem and payoff. Treat repeated interview themes as qualitative evidence, not population-wide counts.

Three survey questions that produce analysable answers:

Which app did you last use to find a new podcast? (Apple Podcasts, Spotify, YouTube, other)
What did you type into the search box the last time you wanted an episode on this topic?
How often do you listen while travelling to work? (daily, weekly, monthly, less often)

An interview script built on past behaviour:

"Walk me through the last time you looked for a podcast on this topic."
"What did you type, and into which app?"
"What did you reject, and why?"
"What made you press play on the episode you chose?"

Keep every question in the past tense. Hypothetical questions invite polite agreement, which produces unusable data.

Research operations, ethics and repeatable process

Listener email addresses, survey answers and interview recordings are personal data. The Data Protection Act 2018 and UK GDPR require a lawful basis, a stated retention period and a route for access requests. The ICO publishes guidance on each.

Decide purpose, storage and deletion before the first question, and write it into the consent wording.

Run the programme as a cycle: plan, collect, analyse, apply. Each stage needs an owner and a date. Without dates, collection drifts and analysis never starts.

Keep a simple keyword map of listener phrases, candidate queries, episode titles and show-note openings. Review it quarterly, recheck whether each query still fits, and refresh older titles rather than replacing whole episodes.

Common questions

How much research does a small podcast need?

One survey and eight to fifteen interviews a year, plus quarterly checks on your own analytics. That is enough to name episodes and rewrite show notes.

Where does UK GDPR bite when I collect listener data?

Recordings, email addresses and survey answers are all personal data. You need a lawful basis and a stated retention period, plus a route for access requests, documented before collection starts.

How do I know when to stop collecting?

Stop when three consecutive interviews produce no new themes. More conversations after that add cost without adding decisions.

Can I reuse last year's keywords?

Yes, if you recheck the query and the title still answers it. Refresh the phrasing rather than the whole episode.