Why You Can't Track Podcast ROI the Way You Track Everything Else
If you've ever sat down to figure out what your podcast is actually doing for your business, you've probably had this experience at least one time.
You open your analytics, you go looking for the podcast, and it isn't there..
Your traffic shows up as "direct." Your conversions get credited to whatever the person clicked last, which is usually a Google search for your own name. And meanwhile you know for a fact that three clients this quarter mentioned the show on their very first call.
So what's going on? Is that analytics platforms are structurally incapable of seeing podcasts. Once you understand why, you can stop chasing a number that's never going to exist.
Why the tracking breaks
There are three things working against you here, and they're all happening at the same time.
1. There's no click
Every measurement system you use was built for the web, where somebody clicks a link and carries a little trail along with them. A referrer, a UTM tag, a cookie. That trail is the whole reason your dashboard can tell you "this person came from Instagram."
Podcasts don't work like that. Your episode is a file that gets delivered through an RSS feed to an app, and then it plays there. There's no click on your ad. There's no referrer getting passed along. Nothing at all connects the moment somebody hears you to the moment they turn up on your website. It's not that the chain breaks. There was never a chain.
And this isn't unique to podcasting, either. It's a whole category of traffic that analytics quietly mislabels. SparkToro ran a study on it and found that visits arriving from TikTok, Slack, Discord, Mastodon and WhatsApp got labeled as "direct" essentially across the board. Not undercounted. Invisible. If your analytics can't see a link somebody clicked inside a group chat, it definitely can't see somebody who heard you in their car.
2. The listening moment is the least clickable moment in media
This is the part people really underestimate. Your listener is driving somewhere. Walking the dog. Doing the dishes. At the gym. On a train.
That's the whole magic of the format, honestly. You get somebody's attention during time they can't give to anything else. But it also means they physically cannot act on what you're saying while they're listening. Their hands are busy and their phone is in their pocket.
So if the action happens at all, it happens later. Hours later, sometimes weeks later, on a totally different device, after they've half forgotten where they heard about you in the first place. Usually they get there by searching your name. And by that point your podcast is completely invisible to every system that's watching.
3. Podcast consideration windows are long
If somebody's buying a $20 impulse product, sure, they might act the same day. But for a $12,000 trip, or a coaching package, or an ongoing therapeutic relationship? Nobody acts the same day. Nobody should.
Relationship based businesses have these long quiet consideration periods where somebody listens for four months, tells absolutely nobody they're doing it, and then books. Most attribution tools have a 30 day window at best. Your actual sales cycle just laughs at that.
Put all three of those together and you get the real problem, which is that last click attribution systematically undervalues anything that builds trust slowly. It's not that your podcast isn't working. It's that the measurement tool was designed to credit whatever happened last, and your podcast's whole job is to be the thing that happened first.
And the size of that gap is genuinely startling once somebody measures it. A twelve month study from Refine Labs compared what attribution software reported against what customers said about themselves, and found roughly a 90% measurement gap for dark social channels, a category that explicitly includes podcasts and word of mouth. Ninety percent. That's the hole your show is falling into.
What to do instead: stack imperfect signals
You're not going to find one clean number, so I'd stop looking for it. What actually works is combining a handful of imperfect measurements that add up to something true.
→ Ask people directly, because this is the highest value thing you can do
Put a "How did you hear about us?" field on your booking form, your intake form, or your contact page.
Make it free text instead of a dropdown. Dropdowns force people into your categories and you lose all the specificity. Somebody writing "your podcast, the episode about visa timing" tells you infinitely more than somebody selecting "Other."
There's a data problem with dropdowns too. Ruler Analytics looked at this and found that among leads who selected the first option in a dropdown, nearly half of those answers were inaccurate. People pick whatever's at the top, or whatever's closest to true, and you end up with tidy data that's quietly wrong.
I know free text feels almost insultingly low tech. It also outperforms your entire analytics stack when it comes to podcast attribution, because it's asking the only system that actually knows the answer, which is the person.
Keep it optional so it isn't adding friction, and then read the answers once a month. Patterns show up fast.
→ Ask again on the call
For relationship businesses this is usually where the real truth comes out. Somebody will write "internet" on a form and then ten minutes into a discovery call they'll say "oh, I've actually been listening to your show since last spring."
That one sentence is worth more than any dashboard. Start writing those down. A simple tally is plenty.
→ Use URLs and codes people can actually say out loud
If you're saying something out loud, it has to survive being remembered in a car.
So keep it short, memorable, and easy to type. Something like yoursite.com/start, not a UTM string with four parameters hanging off it. And give each show you appear on its own simple path, like yoursite.com/thatshow, so any traffic that lands there has exactly one possible origin.
Same idea with discount or reference codes. Make them easy to say and easy to spell.
→ Build dedicated landing pages
A page that exists only for one specific podcast or appearance or campaign is self attributing. Nobody stumbles onto it by accident, so every single visit means something.
This is about as close as podcasting gets to clean tracking, and it takes you an afternoon to set up.
→ Watch for lift instead of clicks
Rather than asking who clicked, ask what changed.
Pick yourself a baseline first. Your normal weekly direct traffic, your branded search volume, form fills, DMs, whatever you've got. Then watch what happens in the days right after an episode drops or a guest appearance goes live. If your direct traffic and searches for your name reliably bump in the 48 hours after you publish, that's your podcast. It's correlation rather than proof, sure, but across enough episodes the pattern gets pretty hard to argue with.
Branded search volume is especially good for this. People who hear you somewhere and get curious go type your name into Google. That's your podcast showing up in disguise.
→ Use podcast native analytics for the top of the funnel
Your hosting platform will tell you downloads, consumption rate, and which platforms your listeners are on. Prefix services like Podtrac add standardized measurement on top of that. And if you're buying ads at real scale, pixel based attribution exists and works reasonably well for bigger campaigns.
Just be clear with yourself about what these are measuring, which is reach and engagement, not revenue. Consumption rate is the underrated one. Are people actually finishing? A small audience that finishes episodes is worth so much more than a big one that bails at minute three.
Reading the signals together
No single one of these is conclusive on its own. Together, though, they get you to a real answer.
Say a quarter of your discovery calls mention the show, your direct traffic bumps every Tuesday when episodes drop, branded search is climbing, and your /start page is getting steady traffic you can't explain any other way. At that point you know exactly what your podcast is doing. You just can't put a single tidy number on it, and that's genuinely fine.
The urge to demand one clean figure comes from working with channels where one actually exists. There isn't one here, and pretending otherwise is how people end up killing shows that were working.
The reframe that matters most
Here's what I'd really want you to take away from all of this.
For most relationship based businesses, your podcast isn't the lead source. It's the reason the lead already trusted you by the time they got to you.
Somebody listens for six months and never clicks a single thing. Then a friend of theirs mentions needing what you do, and they say "oh, you should talk to her." And that referral converts in one call instead of three, because the trust was already built.
Your dashboard writes that down as a referral. Which isn't wrong exactly, but the podcast is why the referral existed at all and why it closed so easily.
I get into this more in my post on why starting a podcast is still one of the smartest moves a business owner can make, but the short version is that prospects who engaged with a podcast before a sales conversation move through the pipeline noticeably faster and close at higher rates. That speed is the podcast doing its job. It just happens to be a kind of work that no attribution model is built to credit.
It's the same reason nobody asks what their reputation's click through rate is. Some assets don't generate clicks. They just make everything else convert better. Podcasts sit right in that category, which is why last click measurement is always going to make them look worse than they are.
Where to start
Do these three things this month:
Add a free text "how did you hear about us?" field to your booking or contact form.
Start noting it on discovery calls any time somebody mentions the show, even casually. A column in a spreadsheet is enough.
Set up one memorable URL you can say out loud during your show or as a guest, and then actually use it consistently.
Give it a quarter and you'll know considerably more than your dashboard has ever told you.
And if you want the workflow behind a show that's worth measuring, from planning all the way through promotion, it's in my free Podcast Production Process Guide.
Your podcast is probably working better than your analytics can see. You just have to ask a different question, and sometimes you have to ask an actual person.
— Erin Beavers
The Intuitive Beavers Agency, Podcast Producer
