Why Founders Don't Understand Their Growth Channels (And Keep Paying For It)

Somewhere around Series A, a pattern emerges. The founder is making marketing decisions based on a mix of gut feel, anecdotal feedback from customers, and whatever number the agency is reporting back on Mondays. Everyone is busy. The data is technically there — UTMs, Google Analytics, some spreadsheet the growth hire put together — but nobody is quite sure if the numbers are right, or how to act on them anyway.

This isn't a story about bad founders or bad teams. It's a story about attribution — and how the tools most companies use to measure marketing effectiveness were designed for a different era of digital advertising.

---

The data-rich, insight-poor problem

Modern marketing stacks generate enormous amounts of data. Session counts, scroll depth, time on page, email open rates, UTM-tagged conversions, funnel drop-off by cohort. Teams drown in numbers.

And yet, in the room where budget decisions are made, the honest answer to "which channel is actually working?" is usually some variation of "we think it's probably podcasts, but the numbers are hard to read."

Why? Because attribution is genuinely hard — especially when your acquisition strategy includes channels that don't produce clean digital fingerprints. Podcast ads are the clearest example. A listener hears an ad, converts a week later, and the only connection between the ad and the sale is a promo code used at checkout or a vague "heard it on a podcast" answer in a post-purchase survey. That's not nothing. But it doesn't show up in your Shopify dashboard as "attributed to The Leverage Podcast."

---

How founders actually make podcast budget decisions

Let's be honest about the process. In most companies, the podcast budget decision goes one of two ways.

Version 1 (optimistic): "We sponsored three episodes with James and we noticed signups went up that week. It seems to be working. Let's renew."

Version 2 (pessimistic): "We gave them a tracking link and we only got 40 clicks. That's £100 per click. Let's cut it."

Both of these are wrong, just in opposite directions. Version 1 is survivorship bias and correlation masquerading as causation. Version 2 is click-based attribution for a channel where most converters don't click tracking links.

The founder who makes Version 1 decisions keeps paying for campaigns that might be underperforming. The one who makes Version 2 decisions kills campaigns that are actually profitable, because the conversions are flowing through channels they're not measuring.

---

The vanity metric trap in creator advertising

There's a specific version of this problem that's unique to creator advertising: optimising for the metric that's most visible, not the metric that matters.

In podcast advertising, the most visible metric is usually download numbers or audience size. Sponsorship pricing is typically pegged to this — CPM rates based on episode downloads. It's the metric the platform sells you on, it's what agencies report, and it's what most brands focus on when comparing podcasts.

But downloads have no direct relationship to conversions. A podcast with 50,000 downloads per episode might have a highly passive, disengaged audience that rarely acts on ad reads. A podcast with 8,000 downloads might have a hyper-engaged niche audience where the host has genuine authority over purchase decisions — and a 3x conversion rate on ad mentions.

If you're making budget decisions based on CPM rates and download counts, you're optimising for the wrong thing. The relevant metric is cost per attributed conversion, or cost per attributed pound of revenue. You cannot calculate either of those without proper attribution.

---

The specific failure modes

Here are the attribution failures we see most often, and what they cost:

No promo code tracking. The creator reads a discount code on every episode. You're tracking clicks but not promo code redemptions. You're seeing 60 conversions when the actual number is 180. You don't renew. You leave a profitable channel behind.

No vanity path. The creator mentions a custom URL verbally. You're only tracking the show notes link. Most listeners who convert type the URL directly — your analytics shows "direct" traffic. You can't connect those sessions to the campaign. The campaign looks like it barely worked.

No post-purchase survey. 20% of your new customers this quarter came from podcasts. None of your attribution tools captured it. You think your paid search is working much better than it is. You shift budget away from podcasts and toward search. CAC rises.

Attribution window too short. You're using a 7-day attribution window. Podcast listeners often convert 2–4 weeks after first exposure. You're systematically undercounting conversions from every campaign.

---

What good creator advertising measurement looks like

The standard for creator advertising attribution should be:

1. Every campaign has a unique tracking link — not a generic UTM, but a dedicated link tied to a specific creator and campaign
2. Every campaign has a vanity path if the creator is reading a custom URL
3. Every campaign has a promo code if the creator is offering a discount
4. Every campaign's results include survey data — the post-purchase survey answer is the ground truth that validates everything else
5. Results are viewed across all four signals combined, not in isolation
6. You can switch between attribution models to understand first-touch, last-touch, and linear impact separately

If you're running 10 creator campaigns and you can tell me the cost per attributed conversion and the attributed revenue for each one — you're doing this right. If you can't, you're flying blind.

---

The compounding cost of bad attribution

Here's the thing about attribution failures: the cost isn't obvious in the moment. It compounds quietly over time.

You cut a profitable podcast because the click data looked bad. You renew an underperforming one because the downloads are impressive and you had a good week that might have been seasonal. You can't make the case to your board for a bigger creator budget because you can't show clear ROI. Meanwhile, a competitor who's got proper attribution is doubling down on the channels that work, cutting the ones that don't, and building a creator advertising flywheel you can't compete with.

The competitive advantage in creator advertising right now isn't about who has access to the best podcasts or the biggest YouTubers. It's about who can measure ROI clearly enough to outbid their competitors with confidence on the campaigns that actually work.

---

Getting started without a 6-month data project

None of this requires a complex data engineering project. The minimum viable attribution setup for creator advertising is:

• A platform that creates per-campaign tracking links (not just UTMs)
• Vanity path routing (a simple redirect, configurable without a developer)
• Promo code ingestion via a basic API or JavaScript tracker
• A one-question post-purchase survey fed into the attribution engine

That's the four signals. It takes a few hours to set up the first time. After that, it runs automatically for every campaign you add.

The founders who understand their growth channels aren't smarter than the ones who don't. They're just using tools that were designed for the channels they're actually running on.

---

Castlytics gives you all four attribution signals, per-campaign ROI, and attribution model switching in one dashboard. Start free — your first campaign is live in under 10 minutes.