How to run a quarterly podcast review

A quarterly podcast review is a structured look at one quarter of releases, using the same measurement window for every episode. It should tell you what changed, what probably caused it, and what you will test next. It should not be a slide deck full of lifetime totals.

You can run the review in a spreadsheet and a one-page decision note. Compare episodes fairly and refuse to add metrics that describe different events.

Before your quarterly podcast review

Gather the people who can act on the result: the host, producer, editor, and whoever owns distribution or sponsorships. One person can run the review for an independent show, but write down the decisions as if somebody else must execute them.

You need access to four data sources:

1. Your host or prefix analytics for downloads and delivery by episode.
2. Apple Podcasts Connect for Apple-only listeners, plays, and consumption.
3. Spotify for Creators for Spotify-only performance and retention.
4. Your business records for signups, leads, memberships, or sponsor results.

These sources overlap, so keep them in separate columns. The difference between podcast downloads and listens explains why a file request, a play, and listening time cannot be collapsed into one audience total.

Choose the quarter and a fixed episode-age window before exporting anything. For example, compare every episode after its first 30 days, not the newest episode at day 6 against an older episode at day 88. Apple Podcasts Connect has a Performance view built around this idea. Apple documents comparisons at 7, 14, 30, or 60 days after release, with median, average, or top episode as the baseline.

Step 1: build one row per episode

Open each analytics dashboard, set the date control to the quarter, and export or copy the episode-level results. Your working sheet should have one row per episode and a source in every metric label.

Start with these columns:

Do not paste a lifetime download figure into the fixed-age column. It rewards older episodes for existing longer. If your provider cannot show episode age directly, calculate the cutoff date from the release date and pull only data through that date.

Record missing data as blank, not zero. Zero says the event did not happen. Blank says you do not know.

Step 2: establish the normal episode

Sort the fixed-age download column and calculate the median. The median is the middle episode after sorting, which makes it less vulnerable than the mean to one unusually large or small release.

Apple's Performance view offers both median and average baselines. Use the median as your operating baseline, then keep the average nearby as a check. A wide gap between them tells you that the quarter contains an outlier worth inspecting.

Next, calculate the difference from the median for each episode:

episode downloads at fixed age - quarterly median

You do not need to turn every difference into a percentage. The raw gap is often easier to discuss, especially for a smaller show. Mark the top two and bottom two episodes, then move on. A quarterly review should explain the edges, not narrate every row.

Step 3: separate reach from consumption

Reach asks whether people arrived. Consumption asks what they did after pressing play. Review them in separate sections.

For reach, look at fixed-age downloads, platform listeners, and the distribution attached to each release. For consumption, look at average consumption and the retention curve. Apple's analytics defines average consumption as the average share of an episode played per device, and notes that it can exceed 100% when somebody listens more than once. That makes it useful inside Apple, but not interchangeable with host downloads.

Inspect the same moments across the outliers:

• The opening minute, where a long setup or housekeeping block can lose people.
• The handoff into the main conversation.
• Mid-roll positions.
• The final segment and call to action.

The drop-off point guide shows how to distinguish a sharp exit from a normal decline. Write what happens at the timestamp before proposing a fix. "Retention fell at 12:40 during a four-minute sponsor and housekeeping block" gives the editor something to change. "Engagement was weak" does not.

Step 4: inspect discovery inputs

Add the promotion attached to each outlier. Did the guest mail their list? Did you publish a clip? Was the title tied to a search query? Did the newsletter go out late? These observations narrow the next test without proving what caused the result.

Look for repeated combinations. If three above-median episodes shared a topic but used different formats, the topic deserves another release. If interviews only rose when the guest promoted them, the booking process needs a distribution requirement. If a strong consumption curve came with weak reach, keep the episode structure and change the packaging or distribution.

This is where reading podcast analytics in the right order helps. Start with delivery, then discovery, then consumption, then the business result. Reversing that order creates stories the data cannot support.

Step 5: connect the quarter to business outcomes

Downloads alone cannot tell you whether the show produced a useful result. Add the outcome your show is meant to create, and keep its attribution method visible.

For a membership show, that might be trials or paid members from a tagged link. For a sales-led show, it might be qualified calls that named an episode. For a sponsored show, it might be delivered impressions, promo-code use, or visits to a vanity URL.

Do not claim that every conversion after an episode came from the episode. Record the evidence you have and the attribution method used. The podcast ad performance guide covers the difference between delivery, response, and business outcome.

A useful outcome table is short:

Step 6: turn observations into tests

End with no more than three tests for the next quarter. Each test needs a change, an owner, a date, and a measure.

Use this format:

Observation: Interview episodes with guest newsletters reached more people at day 30.
Change: Add one confirmed newsletter placement to the guest agreement.
Owner: Producer.
Due: Before the next interview records.
Measure: Day-30 downloads compared with the quarterly median.

A second test might shorten the opening before the main segment. A third might repeat a topic that produced qualified leads. Avoid broad assignments such as "improve promotion" or "make episodes more engaging." Nobody can ship those.

Step 7: verify the review

Before you close the sheet, check five things:

• Every episode uses the same age window.
• Every metric names its source.
• Blanks and zeroes have different meanings.
• Apple, Spotify, and download figures have not been added together.
• Every decision has an owner and a review date.

Then put the next review on the calendar and freeze a copy of the sheet. When the next quarter ends, compare method to method before comparing result to result. A changed filter, window, or metric definition can create an apparent trend that came from measurement rather than listeners.

The IAB podcast measurement guide explains how server-log measurement and filtering shape download figures. IAB Tech Lab's current guidelines page also notes that podcast measurement is based on server logs, unlike media that maintains an open connection while someone consumes it.

If you need a consistent cross-app analytics layer for the next review, start tracking your show with Podder. Install the prefix now so the next quarter has a clean baseline.