Consider a worked example: £50,000 invested in podcast sponsorships, 23 promo-code conversions, and repeated sales notes saying "I found you on a podcast." The evidence does not line up.
It doesn't add up because the measurement is wrong. Not slightly wrong. Fundamentally, structurally wrong — in the same way that measuring the effectiveness of a billboard by counting how many people photograph it would be wrong. You'd conclude billboards don't work. You'd be measuring the wrong thing entirely.
Podcast advertising creates a measurement problem because standard attribution approaches were designed for clickable, cookie-based media. Podcasts are neither. Promo-code data can still be useful, but it should not be treated as a complete estimate of podcast influence.
The Promo Code Problem
Promo codes are a common podcast measurement method. The host reads your ad, mentions a code — "use PODCAST20 for 20% off" — and you count how many people use it. That is simple and clean, but it can miss exposure that converts through another path.
Promo-code redemptions are not a defensible estimate of all podcast-influenced purchases. They measure one observable action and omit listeners who remember the brand, search later, or involve another stakeholder in the purchase.
The reasons are straightforward. A listener may forget the code, purchase later, search for the company by name, or find a different offer. In B2B, the person who heard the ad may not be the person who completes the purchase. The decision-maker can hear the podcast while procurement handles the transaction, leaving the promo code unused between them.
You're measuring the shadow of a shadow, and drawing conclusions about the sun.
Why Pixel-Based Attribution Completely Misses Podcasts
If promo codes don't work, surely pixel-based attribution — the UTM parameters, the cookies, the multi-touch attribution platforms — can pick up the slack? They cannot. And the reason is architectural, not technical.
Pixel-based attribution tracks clicks. Podcasts don't produce clicks. There is no link to click when you're listening on your morning commute. There is no cookie to drop when the medium is audio playing through earbuds. The entire measurement infrastructure of digital marketing was built for a world where the first interaction is a click. Podcasts operate in a world where the first interaction is a thought.
This is not an edge case. This is the typical journey. The listener hears your name, files it somewhere in memory, and weeks later — when a relevant problem surfaces — retrieves it as a Google search. Your attribution platform sees a Google conversion. Your podcast attribution sees nothing. And you conclude the £50,000 you spent on podcasts generated 23 conversions when it actually generated hundreds.
The Solution: Self-Reported Attribution for Podcasts
The fix is disarmingly simple. Ask buyers how they found you. Not with a dropdown menu — those are leading and incomplete. With an open text field. "How did you first hear about us?" Let them answer in their own words.
Two common objections are that people will not remember or that the data will be messy. Both are valid limitations, but neither makes the question useless. Reviewed self-reported answers can add evidence that attribution platforms do not capture, even when the data needs interpretation.
The practical challenge is categorisation. People describe the same podcast in a dozen different ways. This is where AI earns its keep — not generating content, but classifying freeform human responses into structured attribution data.
Six different descriptions. Two distinct categories. Without AI classification, this data sits in a spreadsheet as six unstructured strings nobody analyses. With it, you get a clear signal: podcasts are driving pipeline — and you can see which ones.
Self-reported attribution is not a replacement for pixel-based tracking. It is a complementary source for influences that pixel-based tracking may not observe. For podcasts, combine it with campaign timing, branded-search trends, cohort quality and revenue evidence.
How To Actually Calculate Podcast ROI
Once you have self-reported attribution data, the ROI calculation becomes straightforward. Here's the method, step by step.
Count podcast-attributed conversions from self-reported data
15% of new customers mention podcasts
Apply that ratio to total conversions
500 total conversions × 15% = 75 podcast-attributed
Calculate attributed revenue
75 × £10,000 ACV = £750,000
Compare to podcast spend
£750,000 ÷ £50,000 spend = 15x ROI
In this worked example, promo-code measurement and the expanded attribution estimate produce opposite conclusions. That is a reason to examine the inputs and uncertainty, not to assume either model is automatically correct.
Same illustrative spend and revenue inputs, measured two ways. Replace every input with your own reviewed response, conversion and revenue data before making a budget decision.
Your result may be smaller, larger or absent. The purpose of self-reported attribution is to test for missing podcast influence and reconcile it with revenue evidence, not to guarantee a particular share of pipeline.
Better podcast measurement does not require assuming the channel worked. It requires one additional source of evidence — what customers remember — reconciled with campaign, cohort and revenue data.
Frequently Asked Questions
Is podcast advertising actually effective for B2B?
It can be, but performance varies by audience, show, message, frequency, offer and sales cycle. Podcast exposure often happens away from a clickable session, so evaluate it with customer-reported attribution, branded-search trends, cohort quality and revenue evidence rather than promo codes alone.
How do you measure podcast ROI without promo codes?
Add a self-reported attribution question — an open text field — to your sign-up flow, demo request form, or post-purchase survey. Ask "How did you first hear about us?" and let people answer in their own words. Then classify the responses and compare podcast-mentioning cohorts with your recorded revenue data.
Why do promo codes undercount podcast conversions?
People may forget a code, search for the company later, use a different offer, or pass the purchase to another stakeholder. A promo code therefore measures code redemption, not the full population influenced by a podcast.
How long does podcast advertising take to show results?
The delay depends on your buying cycle and the listener's timing. Measure over a window long enough to cover the normal path from first awareness to revenue, and compare several cohorts rather than applying a universal podcast-conversion timetable.
What's the best way to track podcast-influenced revenue?
Combine three approaches. First, self-reported attribution on every conversion point. Second, branded search lift analysis — measure the increase in branded search volume during and after podcast campaigns. Third, cohort analysis — compare close rates and deal sizes for customers who mention podcasts versus those who don't. Together, these give you a comprehensive picture that no single pixel-based tool can provide.
James Kevan is the co-founder of First Signals, where he helps B2B companies understand what's actually driving their pipeline. If your attribution dashboard and your sales team are telling different stories, the AI Opportunity Audit starts by uncovering the channels you're undervaluing — before recommending any changes.
From the same series: Islands. Good tools. No bridges. · Your Business Isn't Broken. Your Processes Are. · The AI Brain Freeze · The Quiet Businesses.
© 2026 James Kevan / firstsignals.ai. Share freely with attribution.