# Attribution for Podcast and Radio Ad Spend

Source: https://attributionplaybook.com/blog/2026-10-14-attribution-podcast-radio-ad-spend.html

October 14, 2026

*Audio ads reach people offline and convert online, which breaks channel level attribution. How to measure podcast and radio spend without pretending clicks tell the story.*

Podcast and radio spend breaks the attribution model most stores use, because the whole point of audio is that nobody clicks. A listener hears a host read, then visits the store hours later by typing the URL, searching the brand, or tapping a link in the show notes. Last click attribution sees an organic search or a direct visit and hands audio no credit at all.

The usual response is to underinvest, overinvest, or argue about it forever. None of those are necessary. Audio is measurable, but the measurement looks different from paid search: it leans on structured tests, codes, and vanity URLs rather than on click paths.

## Give audio its own trackable mechanics

Unique offer codes and dedicated landing pages per show are the backbone. A listener who types a show specific code is identified regardless of the path they took to the store. Keep the codes genuinely distinct by show and by flight, or the readout blurs into a mush of overlapping promos and the test tells you nothing.

Vanity URLs work for podcast show notes, where a tap is possible, but do not expect most listeners to click. Design the measurement around the code being spoken and remembered: short, easy to spell, and tied to the show name. The code is the attribution.

## Run audio like a test, not a channel

Audio attribution starts with a clean baseline. Launch a show, watch branded search, direct traffic, and code usage for the flight window, and compare against a control period or a matched market. Geographic holds work well for radio: run in some markets, hold others, and compare the lift in branded search and code redemptions.

Expect delay. Audio response often builds over weeks as episodes accumulate listeners and a host read airs repeatedly. A one week flight is usually too short to judge. Plan four to eight weeks, keep the creative and offer stable, and resist the urge to relaunch the offer mid flight, which contaminates the readout.

## Read the blended picture, not the click table

The click based report will say audio did nothing. The store will say otherwise. Put the code redemptions, vanity page visits, and lift in branded search and direct traffic side by side, and judge audio on the total. This is where blended ROAS earns its keep: audio is a brand building channel with direct response mechanics, and it should be measured as a contributor to the whole.

Watch for contamination. A big sale or a viral moment during an audio flight will inflate the numbers, and a host going off script with a different code can strand redemptions in the wrong bucket. Note every anomaly in the readout so the next flight plan is built on the signal, not the noise.

## When audio is not working

Kill criteria should be set before the flight: a minimum code redemption rate, a minimum branded search lift, or a payback threshold on the spend. Audio fails most often from creative, an unread sounding read, or from buying shows whose audience does not buy things online. Both are fixable with a different show and a tighter read, not with a bigger budget on the same plan.

## Frequently asked questions

### Why does last click attribution fail for podcast ads?

Because listeners rarely click. They hear a host read and visit the store later by typing the URL, searching the brand, or using a show link, which last click systems credit to direct, organic search, or the link itself, leaving the podcast with zero recorded conversions.

### How do you attribute a podcast campaign in practice?

With unique offer codes and dedicated landing pages per show, plus lift analysis on branded search and direct traffic during the flight. Codes identify the listener regardless of path, and the lift read captures spillover that never used a code.

### How long should a podcast test flight run?

Usually four to eight weeks. Audio response builds as episodes accumulate and host reads repeat, and short flights mostly measure noise. Keep the creative and the offer stable for the whole flight so the readout is interpretable.

### What is the most common reason audio spend looks like it failed?

Judging it on click based reports, which are structurally blind to audio. The second most common reason is a contaminated test: changing the offer mid flight, overlapping promos, or a sale that makes the lift unreadable.

[Home](../index.md) / [Blog](index.md)