Online Radio Advertising: Plan and Evaluate a Streaming-Audio Campaign - Yenra

Choose the listening environment

On a small screen, scroll the table sideways to read all columns.

A podcast is commonly consumed as an on-demand episode rather than a continuously programmed radio stream. Buying a podcast audience does not automatically buy an endorsement from its host. Ask whether the creative is a separately produced spot, an announcer read, or an actual testimonial; each creates different expectations.

Specify whether reporting identifies people, accounts, devices, downloads, or ad deliveries. Do not add these counts as interchangeable units. The IAB Tech Lab podcast measurement guidance explains why server-log-based download and ad-delivery measurement differs from media with an ongoing client connection. A download is not direct evidence that the full ad was heard.

Write one message that survives a screenless moment

Give the listener a recognizable advertiser, one useful idea, and one next step. Say the brand early enough to connect it with the offer. Avoid a chain of phone numbers, unfamiliar abbreviations, multiple prices, and qualifications spoken too quickly to understand.

Write the destination as it will be spoken, then test whether someone hearing it once can repeat it. A companion banner can help listeners who have a screen open, but the spoken message should still make sense when the phone is in a pocket. Do not tell someone to tap a banner in an inventory format that cannot show one.

Download the sample script and production brief (plain text). It separates spoken copy from production notes and includes a checklist for the offer, intended audience, destination, rights, and review owner. Replace the fictional details and verify every claim before recording.

A useful rehearsal test is to play the draft to someone unfamiliar with the campaign and ask what it offered and what they should do next. If they cannot answer, simplify the message before adjusting the music or adding a stronger sales phrase.

Prepare the files for the actual buying route

Specifications vary by platform, placement, and buying arrangement. As checked September 6, 2026, Spotify's U.S. audio-ad specifications list a 30-second maximum for standard music ads and streaming-ad-insertion podcast ads. They recommend no more than 65 words for a 30-second ad. Longer music formats are limited to particular direct campaigns and markets.

The same page lists MP3, WAV, or OGG, stereo, a 44.1 kHz sample rate, and a loudness target of −16 LUFS with a ±1.5 LUFS tolerance. For encoded delivery, it lists a 192–320 kbps bit-rate range. Follow the requirements for the selected format and placement; a WAV master and a compressed delivery file do not use bit rate in the same way.

Export, then check the exported file rather than only the editing session. Confirm exact duration, intelligibility, peaks, channel behavior, and the absence of clipped starts or endings. Keep music below the speech and listen through headphones and an ordinary small speaker. A louder mix is not necessarily a clearer one.

Prepare any required companion image, advertiser label, and HTTPS destination for that particular format. Keep the visual message consistent with the spoken offer, and check that resizing does not remove a necessary qualification.

Clear the voice, music, and reuse permissions

The U.S. Copyright Office's guidance for musicians distinguishes the underlying musical work from its sound recording; they may have different owners and licenses. Permission to stream a song as entertainment should not be treated as permission to put that recording into an advertisement.

Document the rights for the exact use, including media, territory, duration, edits, and renewal. Record voice-talent terms and whether a recording may be adapted or reused. A familiar song, stock track, or synthetic voice should not enter the final file merely because it is easy to download.

AI can help shorten a draft or produce a temporary voice for timing in an approved workflow. Do not imitate a recognizable person's voice without appropriate authorization, and do not imply that a synthetic performance is a real host's endorsement. Verify pronunciation and pacing yourself. If the ad contains a testimonial, use the endorsement campaign guide to plan truthful claims and disclosures.

Measure reach, repetition, and business response separately

Agree the reporting definitions before the buy. Reach and average frequency describe delivery across the reporting system's identified audience; completion describes a playback or delivery event under its rules. None alone establishes that a person listened attentively or took action because of the message.

The measurement example (CSV) records these assumptions. A real report should identify the reach unit and deduplication method. A rate comparison is meaningful only if the purchased impression definitions match.

Use a relevant landing page or response mechanism, but recognize its limits. A remembered URL may produce a later direct visit; a promotion code can be shared beyond the original listener. Platform attribution also depends on matching and reporting windows. When additional sales are the question, seek a credible experiment and report uncertainty rather than equating all attributed activity with lift.

Review placement, repetition, response quality, and creative together. If listeners remember the offer but cannot find its details, fix the destination. If the message is unclear, revise the script. Renewing a buy should follow a documented conclusion, not simply a high completed-ad count.